The ISI UGA marking scheme: the exam that pays you to leave it blank

30 MCQs · +4 correct · 0 wrong · +1 for a blank · /120 · 120 minutes

This is the rule almost every other paper in India does not have: ISI pays you +1 for leaving a question alone. A wrong answer costs nothing directly, but it costs you the +1 you would have banked — so the penalty is real, it is just hidden in the opportunity cost.

On a 4-option question a blind guess is worth 4 × 1/4 = 1 mark — exactly what the blank pays. Pure guessing gains nothing on average. Eliminate a single option and the same question is worth 4 × 1/3 ≈ 1.33: now answering wins. The whole of ISI UGA strategy is that one comparison, and it is why "attempt everything" advice is actively wrong here.

Pattern source, carried verbatim from the constants the app grades with: ISI UGA pattern (2026 sources); UGB is subjective, never auto-graded. That registry records this pattern as researched from secondary 2026 sources and not yet checked against ISI's own admission notice, so treat the marks and the duration on this page as unconfirmed until it has been. The option count per question is not a constant we hold either, so the break-even above is stated for k options generally. UGB, the subjective paper, is proof-based and ViJEEta never auto-grades it — the same refusal we apply to RMO and INMO.

Why this page exists

A marking scheme is one of the few things about an exam that is simply a fact — so it is one of the few things we can put on a page without a caveat. Everything on ViJEEta that is estimated (readiness, mastery, chapter weightage) carries its method, its minimum evidence, and a "—" wherever we have not measured it. How the readiness score works sets out what each number means and when ViJEEta refuses to show one.