Skip to content
ExamTrack Prep

Tools

Percentile and cut-off distance

A score out of 100 is meaningless until you know what everyone else scored. This places a score in the distribution and shows how far it sits from the bar a given shortlist ratio implies.

Prelims shortlists roughly twenty times the vacancy count, so the bar is lower than people expect.

Out of 100

Change this if you have better information

Typically 12 to 22 marks

Approximate is fine

Where that score sits

85.9
Estimated percentile
1,26,683
Approximate rank
+1.08
Std deviations from average

A score of 58 is above the average of 44 by 1.08 standard deviations. Under this assumption about the spread of scores, roughly 85.9 per cent of candidates score at or below it, leaving about 1,26,682 candidates ahead of you out of 9,00,000.

How far is that from getting through?

Boards shortlist a multiple of the vacancies rather than a fixed score. Adjust the ratio to see how the bar moves.

Around 20 for a screening stage, 1 to 3 for a final merit list.

Assuming 40,000 vacancies

8,00,000 places, so the bar is near 28.1 marks

You are about 29.9 marks clear of it.

That is the 11.1th percentile, or the top 88.89 per cent.

Read this before you trust the number

  • This assumes scores follow a bell curve. Real distributions are lumpier, particularly just above a cut-off, and railway papers vary by shift before normalisation.
  • The default average and spread are reasonable working figures, not published data. No board releases the mean or the standard deviation of its candidate pool.
  • Cut-offs are also category-wise and state-wise or board-wise. A single national figure does not exist for either exam.
  • Use it to answer "is my mock score in the right region", not "will I be selected". For the latter, only the official result page counts.

The one thing this is genuinely good for

Candidates consistently misjudge which stage they are preparing for. A prelims shortlist calls roughly twenty candidates per vacancy, which sounds brutal until you notice that with 900,000 applicants and 40,000 vacancies the shortlist is 800,000 places. The prelims bar is low. Almost anybody who prepares seriously clears it.

Mains is where the exam is decided, and the pool there is entirely different: everybody still in it has already cleared a screening stage. A score that put you in the 85th percentile of all applicants might sit in the 40th percentile of that group. Slide the shortlist ratio down to 1 or 2 and watch the required score move. That movement is the real shape of the competition, and it is the reason a prelims-level preparation plan stalls at mains.

How to read a percentile without fooling yourself

  • Your mock test percentile is not your exam percentile. The people taking mock tests on any given platform are a self-selected group, usually more serious than average, so their scores skew high. A 60th percentile in a good mock series can be an 85th percentile in the real pool.
  • Percentile is not percentage. Scoring 70 per cent of the marks and being in the 70th percentile are unrelated statements, and in a hard paper they can be very far apart.
  • Track its direction, not its value. Whether your estimated percentile is 85 or 88 is noise. Whether it has moved from 60 to 85 across two months is the signal, and it is the only thing in this tool that is robust to bad assumptions.

Frequently asked questions

Do IBPS and RRB publish percentiles?

No. Both publish your marks and, in some cases, category-wise cut-off marks, but neither publishes a percentile against the full candidate pool. Percentile talk in exam preparation is a convenience for comparing yourself to a group, not an official figure you will ever receive.

Then what is this calculator for?

Calibration. A mock score of 62 out of 100 means nothing on its own. Knowing that 62 sits roughly one standard deviation above a typical average, which is around the 84th percentile, tells you whether you are competing for a screening shortlist of twenty per vacancy or a final merit list of one per vacancy. Those are very different targets and they need different preparation.

Why does it ask for a standard deviation?

Because a percentile depends on how spread out scores are, not only on the average. In a paper where everyone scores between 40 and 50, a 55 is exceptional. In a paper where scores range from 20 to 90, a 55 is ordinary. The spread is what carries that information, and 12 to 22 marks is the usual range for a 100 mark paper.

How reliable is the estimate?

As a rough placement, reasonable. As a prediction, not at all. It assumes a normal distribution, which real exam scores only approximate, and it uses an average and spread this site cannot know. Treat the output as an order of magnitude: top quarter, top ten per cent, top one per cent.

How is a cut-off actually decided?

By working down the ranked list until the required number of candidates is reached, separately for each category and each state or board. That is why cut-offs move every year and why last year's figure is a guide rather than a target. It also means a cut-off can never be computed before the result exists.

Related reading