NEET Negative Marking: When to Attempt, Skip or Return

If the applicable scheme is +4 for correct and −1 for incorrect, the expected value of an attempt with probability p of being correct is 4p − 1(1 − p), or 5p − 1. It becomes positive above p = 0.20. This is a mathematical threshold, not a command to mark every question: your estimated probability may be badly calibrated, time has value, and option elimination may not create equal remaining probabilities.

Use the equation to understand why informed elimination can be worthwhile, then use your own mock data to set behaviour. Suppose a tag produces 30 attempts with 18 correct and 12 wrong: its net is 72 − 12 = 60 marks. Another tag with 6 correct and 14 wrong nets 24 − 14 = 10, still positive but may consume time better spent elsewhere. Analyse marks and minutes together.

Attempt now when you recognise the method or fact, the work fits your pass, and your selected option survives a quick verification. Skip for the paper when the concept is unknown, the stem is not interpretable, or your remaining choice is unsupported. Return when there is a plausible route but it is long, an answer needs one check, or elimination has narrowed the field without enough confidence for the first pass.

These states prevent two expensive behaviours: fighting a difficult question because time has already been spent, and abandoning a solvable question merely because the first step was not instant. Mark return candidates in a consistent way permitted by your test process. Do not create five symbols that require decoding later. One mark for ‘workable later’ and one for ‘do not revisit’ is enough.

Tag each mock attempt before seeing the answer. A useful scheme is A for direct knowledge or a completed derivation with checks, B for a sound method with one uncertainty, and C for elimination or intuition without a complete solution. Also record a reason code such as fact, calculation, units, option elimination or wording. Confidence recorded after the key is contaminated by hindsight.

Every two or three mocks, calculate accuracy and average time for each tag. If B attempts are accurate but slow, the issue is speed rather than risk. If A contains repeated wrong answers, overconfidence or verification is the priority. If C is unproductive, tighten the conditions under which you use elimination. Your decision rule should change only from a meaningful sample, not one lucky paper.

Elimination is strongest when based on independent constraints. In Physics, test units, sign, limiting cases, graph shape and order of magnitude. In Chemistry, test conservation, chemical feasibility, periodic trends, reagent role and stated conditions. In Biology, test exact scope, qualifiers and contradictions with the prescribed source. Eliminate an option only when you can state the reason, not because it looks unfamiliar.

When two options remain, pause before upgrading confidence. Ask whether both were evaluated on the same criterion and whether the stem contains NOT, incorrect, except, most or least. If one option is supported and the other merely feels less familiar, you do not have genuine elimination. In review, record false eliminations separately; they can make an apparently rational strategy systematically worse.

In the first pass, collect high-confidence marks and tag return candidates without opening long battles. The second pass is for questions with a known route, productive elimination or a checkable uncertainty. A final pass should prioritise answer-transfer verification and unresolved high-value candidates, not a frantic sweep of every blank. Set subject and pass budgets from full-length mocks because your speed profile is personal.

Protect a final verification buffer. Check question number alignment, chosen option, and any changed response according to the official response-marking instructions for your exam. Negative marking strategy fails if a sound solution is transferred against the wrong number. Practise the same response-recording rhythm in mocks; exam day is not the place to invent a new batching method.

After each mock, create a table by subject, confidence tag and action. Record count, correct, wrong, net marks and total time. Then inspect counterfactuals: which skipped questions were actually within your knowledge, which attempted questions had no defensible basis, and which return questions became easy after the paper? This shows whether you are too cautious, too aggressive or simply sequencing badly.

Turn the finding into one rule for the next mock. Examples: attempt B-calculation questions only after a unit check; return to lengthy Physics numericals after direct Chemistry items; or do not mark C-fact questions unless two options are rejected with explicit reasons. Compare the next result. A rule that sounds smart but does not improve net marks, time use or response accuracy should be revised.