A second-order effect of the 2026 pattern change is worth stating on its own, because it is quietly making a lot of advice wrong.
With 180 questions at +4, the maximum mark is 720. In 2025, with 200 questions, it was 800. Nothing about the marking scheme itself changed — it is still +4, −1, and 0 — but the denominator did.
What this breaks
Raw score targets. A number like "aim for 550" was calibrated against a paper worth 800 marks. As a fraction that is 68.75%, which on the 2026 paper is roughly 495. Carrying the old number across unchanged sets a target about 11% harder than the one it was meant to express.
The same applies to every "safe score for a government seat" table, every screenshot of a previous year's scorecard used as a benchmark, and every mock series still scoring out of 800.
What this does not break
The qualifying cutoff itself. NBEMS sets it as a percentile — the 50th for the general category, lower for reserved categories and PwD candidates — not as a fixed raw mark. A percentile is a position within the cohort that sat the paper, so it moves with the paper and needs no adjustment.
That distinction is the useful one: percentile targets carry across the change, raw mark targets do not.
The practical version
If you are tracking a raw score from any source published before mid-2026, multiply it by 0.9 before comparing it against a 2026 mock. And prefer percentile to raw marks as your progress measure — it is the only one of the two that survived the change, and it is what your rank is built from anyway.