SSC normalization uses the equipercentile method from the official notice dated 02-06-2025. An exact normalized or safe score needs every shift's marks — published only after the result.
No tool can compute an exact safe score before results
Every number below comes live from the official formula — no invented safe score.
Shift 1 · N = 4 candidates
Step 1 · percentile P = m / N per raw score
| Raw marks | m | Percentile P |
|---|---|---|
| 80 | 4 | 1 |
| 60 | 3 | 0.75 |
| 40 | 2 | 0.5 |
| 20 | 1 | 0.25 |
Shift 2 · N = 5 candidates
Step 1 · percentile P = m / N per raw score
| Raw marks | m | Percentile P |
|---|---|---|
| 90 | 5 | 1 |
| 75 | 4 | 0.8 |
| 60 | 3 | 0.6 |
| 45 | 2 | 0.4 |
| 30 | 1 | 0.2 |
Steps 2–3 · collated percentile column → normalized score
Each shift's raw score at every percentile (actual, interpolated, or boundary), then Z = average across shifts.
| Percentile P | Shift 1 raw | Shift 2 raw | Normalized Z |
|---|---|---|---|
| 1 | 80actual | 90actual | 85 |
| 0.8 | 64interpolated | 75actual | 69.50 |
| 0.75 | 60actual | 71.25interpolated | 65.63 |
| 0.6 | 48interpolated | 60actual | 54 |
| 0.5 | 40actual | 52.50interpolated | 46.25 |
| 0.4 | 32interpolated | 45actual | 38.50 |
| 0.25 | 20actual | 33.75interpolated | 26.88 |
| 0.2 | 20boundary | 30actual | 25 |
Illustrative two-shift example running the exact official formula end to end — check each number by hand. The values follow the method faithfully but are not the notice's own scanned table.
The equipercentile formula needs data that isn't public before SSC releases shift-wise raw marks:
What this page can do
What it will never claim
After results, SSC publishes your normalized marks — use this page to re-check that figure.
Step 1
Inside each shift the Commission ranks candidates by raw marks and converts every raw score into a percentile. This is done separately for every shift, so the difficulty of your particular shift is baked in before shifts are ever compared.
P = m / N N = candidates who actually appeared in the shift. m = candidates in that shift whose raw score is less than or equal to yours. The notice recommends rounding P to 8 decimal places, and 0 ≤ P ≤ 1.
Step 2
All shifts are collated into one table with a single percentile column but a separate raw-score column per shift, sorted from the highest percentile down. Where two shifts have a candidate at the same percentile, their raw scores are treated as equivalent. Where a shift has no candidate at a given percentile, its blank raw score is filled by linear interpolation between the nearest actual scores above and below.
X = x1 + ((x2 − x1) / (p2 − p1)) × (P − p1) X = the blank raw score being filled at percentile P. (p1, x1) = nearest actual point below P. (p2, x2) = nearest actual point above P. This is the equipercentile method named in the notice.
Step 3
For each percentile, the raw scores that correspond to it across every shift — actual where a candidate sat there, interpolated where they did not — are averaged. That average is the normalized score, and it is what feeds final merit and cut-off.
Z = (u1 + u2 + … + ut) / t u1 … ut = the corresponding raw scores across the t shifts at your percentile. t = number of shifts. Z = your normalized score.
SSC uses the equipercentile method from its notice dated 02-06-2025. Step 1: find your percentile in your shift, P = m / N — m is how many candidates in your shift scored at or below you, N is how many appeared. Step 2: put all shifts on one percentile scale, filling any missing raw score by linear interpolation. Step 3: your normalized score is the average of each shift's raw score at your percentile.
No. Any tool that gives you an exact safe score before results is guessing. The formula needs every shift's full marks data (for N and m), and SSC publishes that only with the result. This page works out what can honestly be worked out — and never makes up a safe-score number.
Percentile: P = m / N (rounded to 8 decimal places). To fill a blank raw score X at percentile P: X = x1 + ((x2 - x1) / (p2 - p1)) * (P - p1), where (p1, x1) is the nearest actual point below P and (p2, x2) the nearest above. Normalized score: Z = (u1 + u2 + ... + ut) / t — the average of the matching raw scores across the t shifts.
Any SSC exam held in more than one shift — CGL, CHSL, MTS, GD, CPO and Stenographer among them. The SSC CGL 2025 and 2026 notices both cite the same 02-06-2025 notice, and normalized scores decide the final merit list and cut-off.
It depends on how hard your shift was. A harder-than-average shift usually pushes scores up; an easier one pulls them down. The direction can be guessed, but the exact change can't be known until every shift's marks are published.
No. This page follows the official formula so you can understand and check it yourself. After results, SSC publishes your normalized marks directly — this tool only helps you re-check that figure. SarkariDoc is not an official government website.
SarkariDoc is not an official government website. This tool explains and reproduces SSC's published method; it does not generate an official score.