Every year after results, the same confusion floods GATE forums: "I got 62 marks — why is my score 780?" Because GATE reports two different numbers, and admissions run on the second one. Here is the distinction, minus the mystery.
Marks vs score
- Marks (out of 100): your raw performance on the paper — correct answers minus negative marking.
- GATE Score (out of 1000): a standardised number computed FROM your marks, designed to be comparable across years and, for multi-session papers, across sessions. This is what M.Tech admissions and PSUs actually use, and it is valid for 3 years.
Why normalization exists
High-registration papers (CS is the canonical case) run in multiple sessions with different question papers. Sessions inevitably differ slightly in difficulty, so comparing raw marks across sessions would be unfair. Normalization maps every candidate's marks onto a common scale using the performance statistics of each session (means and toppers' marks), producing normalized marks — which then feed the score formula. Single-session papers skip normalization; raw marks feed the formula directly.
Two practical consequences: you cannot compare raw marks with a friend from a different session and conclude anything; and "easier session" complaints are already priced in by the mechanism.
What the 0-1000 score responds to
The score formula anchors on two reference points — the qualifying cutoff (mapped to 350) and the top performers (mapped toward 900-1000). Between them, your score reflects where you stand relative to the field, not just your absolute marks. This is why identical raw marks can produce different scores in different years: the field moved.
The strategic takeaway: rank-relative performance is the game. Adding 5 marks in a zone where thousands of candidates are bunched moves your score and rank dramatically; the same 5 marks at the top end moves less. For most candidates, the bunching zone is exactly the 25-45 mark band — which is why eliminating silly errors and mastering the high-weight subjects pays disproportionately (the preparation system).
Cutoffs, in perspective
Qualifying cutoffs (typically 25-33 marks depending on paper and category) only make you eligible for a scorecard. Real thresholds are set by what you want: top IIT M.Tech programs and PSU shortlists typically require scores in the 700-850+ range, which in most papers means raw marks well north of 55-60. Set your mock targets from your goal institution's historical closing scores, not the qualifying cutoff.
What to do with all this
Nothing during the exam — normalization is out of your hands. Everything before it: train on real-format mocks scored the real way, watch your percentile trend across mocks (the rank-relative signal), and drill the formats where the field loses marks — MSQ precision and NAT fluency. Myndaq's GATE mocks report exactly that per-format, per-subject breakdown across all six branches, including EE.
Mechanism described per the official GATE scoring documentation; formula constants vary by year and paper. Current as of July 2026.