When an exam runs in multiple shifts of different difficulty, normalization adjusts scores so no one gains or loses from the slot they sat. Your raw score becomes a percentile within your shift, then RRB maps it onto a base shift to produce final normalized marks. This explainer shows the formula, a worked example, and why your result score differs from the answer-key marks.
By Vishal Thakur, Railways & State Recruitment Editor. Published 15 June 2026. Last verified 15 June 2026 against the RRB and SSC normalization methods.
In short
- Normalization adjusts your marks when an exam runs in several shifts of different difficulty, so no one is rewarded or punished for the slot they got.
- Your raw score (correct minus negative marking) is turned into a percentile within your shift — the shift topper gets 100.
- RRB then uses a base-shift formula to convert percentiles into final normalized marks for the merit list.
- Your merit and cut-off are decided on the normalized score, not your raw marks — which is why your "marks" can look different from your result.
If you have taken an SSC or RRB exam and your result score did not match the marks you calculated from the answer key, normalization is almost always the reason. Large exams run across many days and shifts, and no two question papers are equally hard. Normalization is the statistics that level the field. Here is how it actually works, in plain language.
Why normalization exists
When lakhs of candidates sit the same recruitment, the exam cannot be held in one shift — it runs across multiple days and time slots, each with a different question paper. Some papers turn out harder than others. Without a correction, a candidate who drew an easy shift would have an unfair edge over an equally able candidate who drew a tough one. Normalization removes that shift-luck by judging you against the people who sat your shift, then placing everyone on a common scale.
Step 1 — your raw score
First, your raw score is computed from the answer key:
Raw score = (correct answers × marks per question) − (wrong answers × negative marking)
For most RRB papers that is +1 per correct answer and −1/3 per wrong one; unattempted questions score zero. This raw score is the input to normalization, not the final result.
Step 2 — raw marks become a percentile
Within each shift, raw scores are converted to a percentile on a 0–100 scale. A percentile is not a percentage — it is your rank position inside your shift:
Percentile = (number in your shift scoring at or below you ÷ total in your shift) × 100
The topper of every shift gets a percentile of 100, regardless of how hard that shift was. Percentiles are calculated to several decimal places to keep ties rare.
Step 3 — the base-shift formula (RRB)
RRB picks one shift as the base shift (typically the one with the highest average score and full attendance). Percentiles from other shifts are then mapped onto the base shift's scale using an interpolation formula, producing your final normalized marks. The upshot: two candidates with the same ability in different-difficulty shifts end up with comparable normalized scores. The second-stage railway graduate test that uses this and the staff-selection graduate exam both publish results on the normalized scale.
What this means for you
- The score on your result/scorecard is the normalized score, not the raw marks you added up from the answer key — so a gap between the two is normal, not an error.
- Your cut-off and merit position are decided on normalized marks. When you check a previous year's cut-off, that figure is already normalized.
- If your shift was genuinely harder, normalization works in your favour; if it was easier, it pulls your score slightly down. Either way the goal is fairness, not penalty.
- You cannot "game" normalization — the only lever you control is your raw accuracy and attempts. The track you sit also matters; how the NTPC graduate and undergraduate tracks differ explains those separate merit lists, and you can see live openings on the railway jobs hub.
Normalization: हिंदी सारांश
जब कोई भर्ती परीक्षा कई शिफ्टों में होती है और हर शिफ्ट के प्रश्नपत्र की कठिनाई अलग होती है, तब नॉर्मलाइजेशन अंकों को संतुलित करता है ताकि किसी को आसान/कठिन शिफ्ट का अनुचित लाभ या हानि न हो। पहले आपका रॉ स्कोर (सही × अंक − गलत × निगेटिव मार्किंग) निकलता है, फिर उसे आपकी शिफ्ट के भीतर पर्सेंटाइल (0–100) में बदला जाता है — हर शिफ्ट का टॉपर 100 पर्सेंटाइल पाता है। RRB एक बेस शिफ्ट चुनकर इंटरपोलेशन फॉर्मूले से अंतिम नॉर्मलाइज्ड अंक निकालता है, और मेरिट व कट-ऑफ इन्हीं नॉर्मलाइज्ड अंकों पर बनती है। इसलिए आपके रिजल्ट के अंक उत्तर-कुंजी से गिने अंकों से भिन्न हो सकते हैं — यह सामान्य है।