National estimates · the method

How we estimate your CUET percentile and rank.

After a full-length mock, your result shows an estimated CUET percentile and an expected CUET rank for each subject paper, both as a range. This page explains how we work them out, which papers get them and how we check them against real CUET results.

Accountancy · full mock

Example

Estimated CUET percentile

about 94–96

Expected CUET rank

about 9–15 thousand

in a paper taken by about 2.4 lakh candidates

250
marks per subject paper, on the CUET scale: 40+
questions make a paper full length: 40%
of the questions answered, at the least: ±1
standard deviation either side: the width of the range

Step by step

From your mock mark to a CUET percentile.

Every subject paper in the mock is worked out on its own, in four steps.

1Your mark, on the CUET scaleEach subject paper is marked out of 250, the same scale as CUET.
2Adjusted for difficultyA little up or down if that mock was easier or harder than our other mocks in the subject.
3Read as a CUET percentileUsing that CUET year's score–percentile data for the subject.
4Worked into a rankFrom the same percentile and the number of candidates who took the subject that year.

The CUET scale and normalisation

Each subject paper in a mock is marked on the CUET scale, out of 250, so your mark reads the same way a CUET score does.

In CUET, a subject paper is often held in more than one shift. When that happens, NTA may normalise scores so that candidates in different shifts are compared fairly. We don't guess at that step. For each CUET year we use the score–percentile data published or verified for every subject, so the percentile you see follows the method NTA used that year.

Adjusted for the paper's difficulty

Some mocks turn out a little easier or harder than others. If, across everyone who took it, a mock came out easier or harder than our other mocks in that subject, we adjust your mark a little. An easy paper doesn't flatter you, and a tough one doesn't pull you down.

When only a few students have taken a mock, we can't be as sure how hard it was, so the adjustment is scaled down.

Which papers get one

Only papers that look like the real CUET.

An estimate from a ten-question test would mean very little, so national estimates are shown only where they mean something.

Kinds of paper

Full mocks
Previous-year papers
All India Tests
Full-length weekly mocks

Chapter-wise tests, P.Y. Chapter-wise papers and short papers don't get one. They show your rank in the mock, your accuracy and your chapter strengths instead, which tell you more about a test on a few chapters.

Full-length papers onlyThe subject paper has at least 40 questions, so it is close to the real CUET paper.
Attempts that countYou used at least 25% of the time, or answered at least 25% of the questions. A paper submitted earlier than that isn't counted in your reports.
Enough of the paper answeredAt least 40% of the subject's questions answered, so the estimate rests on a real sitting.

Subjects with CUET data

A subject gets a national estimate only once we have its CUET data.

We are adding CUET data subject by subject. Until a subject has it, its papers show no national estimate. You still get your mark, your rank in the mock and your chapter strengths.

Why it's a range

One sitting is not your whole story.

Nobody scores exactly the same twice. A single sitting can come out a few marks above or below your usual level, depending on the paper, the day and how the time went.

So instead of one number, we show a range around the likely value, about 1 standard deviation either side of it. Your likely percentile sits inside that range.

The expected CUET rank comes from the same percentile and the number of candidates who took that subject in that CUET year. A percentile is the share of candidates who scored the same or less, so the rest scored more:

expected rank = 1 + candidates × (100 − percentile) ÷ 100

The best rank in the range comes from the top of the percentile range and the worst from its bottom, so the rank and the percentile always agree.

Example

If a range were 94–96 in a paper taken by 2,40,000 candidates:

Top of the range · percentile 96: 1 + 2,40,000 × 4 ÷ 100 = 9,601
Bottom of the range · percentile 94: 1 + 2,40,000 × 6 ÷ 100 = 14,401

Estimated CUET percentile

about 94–96

Expected CUET rank

about 9–15 thousand

These numbers are made up to show the working. Ranks are shown rounded, so the range reads about 9–15 thousand, with about 12 thousand as the likely rank.

How we check against real results

When CUET results come out, we ask students on their CUETMOCK account: “Got your CUET result?” Students who share their real CUET scores help us check our work.

We compare each real result with the range that student saw after their last full-length mock in the subject. Our staff use what we learn to tune the method for the next CUET.

Percentile among CUETMOCK students

Your result also shows a leaderboard percentile. That one compares you only with CUETMOCK students who took the same mock.

It is a different figure from the estimated CUET percentile, which reads your mark against that CUET year's data for the subject.

An estimate, not a promise.

The CUET paper, the way scores are normalised and the candidates who sit it all change every year. Use your estimate to plan your revision and your college choices, not as a guarantee of your CUET result.

See the mock tests →