Fixed-Paper Percentile vs Randomised Mock Tests: How the Math Works
By Pareeksha · Updated
Short answer: a percentile only means something when everyone was measured on the same questions. On Pareeksha every mock test is a fixed paper, so your percentile and rank are a like-for-like comparison with everyone who sat that exact paper.
Why the same questions matter
Percentile answers one question: “what share of other candidates did I score higher than?” That is only a fair question if the other candidates faced the same test. Suppose two students each score 70 out of 100. If one got a set of easy questions and the other a hard set, those two 70s do not represent the same ability, yet a ranking built on them treats them as equal.
Some mock-test apps assemble a different question set for each candidate from a large pool. That has real benefits for variety, but it means the percentile is an estimate: it blends how well you did with how hard your particular draw happened to be. A fixed paper removes that variable. Same questions, same order, same marking, so the only difference between two scores is performance.
How Pareeksha computes percentile and rank
- You finish a paper. Your score is converted to a share of the paper's maximum marks.
- That share is placed in one of 21 score bands, each covering 5% of the maximum marks.
- Percentile = the share of all attempts on that paper that fell in lower bands than yours.
- Rank = one more than the number of attempts that fell in higher bands.
- Re-attempts are practice only. They are not added to the pool, so repeating a paper cannot inflate anyone's standing.
A worked example
Illustrative numbers, not a real paper. The bands are grouped here for readability; the real histogram keeps all 21 bands of 5%.
| Score band (of max marks) | Attempts |
|---|---|
| Below 40% | 3,200 |
| 40% to 59.9% | 5,800 |
| 60% to 64.9% | 1,300 |
| 65% to 69.9% (you are here) | 900 |
| 70% and above | 800 |
That is 12,000 attempts. Attempts in lower bands: 3,200 + 5,800 + 1,300 = 10,300, so your percentile is 10,300 ÷ 12,000 ≈ 86. Attempts in higher bands: 800, so your rank is 801 of 12,000. Because bands are 5% wide, two candidates in the same band share a percentile; the rank counts only those clearly above.
What a mock-test percentile does not tell you
- It is not your official rank. It compares you with people who attempted that paper on Pareeksha, not with every candidate who applies for the real exam.
- Real exams normalise. Multi-shift exams such as SSC and RRB CBTs adjust scores across shifts of different difficulty. A fixed paper is one shift, so it does not need that adjustment, but it also does not replicate it.
- Small pools are noisy. A paper with a few dozen attempts can move a lot with one strong cohort. Trust the trend across several papers more than any single number.
- Cut-offs come from the notification. Use the percentile to gauge readiness, and the official result for the actual cut-off.
Frequently asked questions
What is a fixed-paper percentile?
A fixed-paper percentile compares your score with everyone who attempted the exact same set of questions. Because the paper is identical, the only thing that differs between two candidates is how they performed, so the percentile measures relative performance rather than luck of the draw.
Why is a percentile from a randomised question pool less reliable?
When each candidate gets a different mix of questions, one person may receive an easier set than another. Their scores are then not measured on the same scale, so a percentile built from those scores mixes ability with question difficulty. It can still be a rough guide, but it is an estimate, not a like-for-like comparison.
How does Pareeksha calculate percentile and rank?
Every finished attempt is placed into a 5%-of-maximum-marks score band for that paper. Your percentile is the share of attempts that landed in lower bands, and your rank is one more than the number of attempts in higher bands. Re-attempts are excluded so practising the same paper again does not distort the pool.
Is a Pareeksha percentile the same as my real SSC or RRB result?
No. It tells you where you stand among the students who attempted that paper on Pareeksha. Official results use the exam body's own scoring and, for multi-shift exams, normalisation across shifts. Treat the percentile as a strong practice signal, and always check the official notice for real cut-offs.
How many attempts are needed before a percentile is meaningful?
The more attempts a paper has, the steadier the percentile. With only a handful of attempts it can swing widely, so Pareeksha shows aggregate statistics such as the average score and cohort percentile only after a minimum number of attempts.
Related: previous year papers · FAQ · pricing