NTA JEE Main 2026: In the middle of ongoing discussions around scores and rankings, the National Testing Agency (NTA) on Tuesday posted a detailed note on X explaining how percentiles, raw marks and normalisation work in JEE Main 2026. The explaination, backed by data from the April session, attempts to address a question many students raise each year, why similar marks can lead to different percentiles.
Wide variation in marks across shifts
According to the data shared, the marks required to reach the 99th percentile varied significantly across the nine shifts conducted between April 2 and 8. In the toughest shift, students needed around 165 marks, while in the easiest, the cutoff rose to 196, a gap of 31 marks out of 300.
A similar trend was seen at other levels too. The difference stood at 27 marks for the 98th percentile and 26 marks for the 97th percentile. Interestingly, only two shifts saw a perfect score of 300. In another shift, the highest score recorded was 285, which still translated to a 100 percentile because no one scored higher in that paper.
The agency pointed out that such variation is not unusual but expected in large-scale exams conducted over multiple days and shifts.
Why percentile, not just marks
Explaining its approach, NTA said no two question papers can be perfectly identical in difficulty, despite multiple rounds of moderation. This makes raw marks alone an unreliable way to compare candidates across shifts.
Instead, the exam uses the percentile system. In simple terms, a percentile reflects how many candidates a student has outperformed within the same shift. For instance, a 99.5 percentile means the candidate performed better than 99.5% of those who wrote that specific paper.
Within a shift, ranking is still based on raw marks, since all candidates face the same paper. The complication arises when results from different shifts have to be combined.
How normalisation works
This is where normalisation comes in. NTA calculates percentiles for each shift separately and then merges them to prepare a common merit list. The idea is straightforward: students who perform equally well relative to their peers, even in different shifts, should be treated equally.
The agency gave a simple example to explain this. If raw marks were used directly, a student scoring 180 in a tough paper could end up ranked lower than someone with the same marks in an easier shift. That, it said, would unfairly reward luck over performance.
“A tested and established method”
NTA emphasised that the system is not new or experimental. It is based on established principles of educational measurement used globally and has been reviewed by expert committees under the Ministry of Education.
The agency also underlined that the method is refined regularly based on data and feedback.
Addressing concerns from candidates comparing scores across shifts, NTA said the percentile is designed to neutralise the impact of paper difficulty. “The shift you were allotted has no bearing on your rank,” the note said, adding that percentile remains the most consistent way to measure performance in such exams.
For students puzzled by why a friend with similar marks may have a different percentile, the message was clear: it is not the marks alone, but how you performed relative to others in your shift that ultimately determines your standing.
