Reasoning Ability — Notes
Core mental model
•Reasoning Ability rewards preparation quality (pattern recognition built through repetition) over raw intelligence — this is the section where consistent daily practice has the most direct, measurable payoff.
•The core puzzle method: read once for shape (no writing) → start from the most definite/positional clue, not the first clue listed → build incrementally, marking uncertain vs. confirmed distinctly → re-verify EVERY clue (including negative ones) against the final arrangement.
•Definite/positional clues should be used first because they fix real positions; negative/exclusionary clues ("not adjacent to") only narrow an already-partially-built set and have nothing to eliminate against if used first.
Topic-specific disciplines
•Syllogism: a conclusion is only valid if it holds in EVERY possible diagram consistent with the statements — checking against just one plausible diagram is the most common source of errors. Practice drawing 2-3 alternative valid diagrams per statement set.
•Blood relations: draw the family tree, never track mentally beyond 3 relations — not a beginner's crutch, a permanently reliable method, since working memory reliably fails under time pressure exactly where a single misremembered relation cascades into every later answer being wrong.
•Direction sense: sketch every problem — "turn right" is state-dependent on current facing, and mental state-tracking across multiple turns is exactly the failure mode a sketch externalizes and eliminates.
•Input-Output: confirm the identified rule against EVERY given transition step before extrapolating to an unshown step — a rule that fits step 1→2 might not hold for 2→3; guessing from partial confirmation is the primary error source.
•Inequality: mixed strict (>) and non-strict (≥) chains need deliberate practice — "either-or" conclusions are a distinct category from checking whether either conclusion individually holds.
•Coding-decoding: practice recent-year pattern variety specifically, not just older fixed-shift patterns — pattern styles genuinely shift year to year, one of the few reasoning topics where staying current with recent exam patterns has direct, specific value.
Multi-variable and Mains-level puzzles
•Complex puzzles (seating + floor + profession) use the same fundamentals-level method — what changes is the tracking structure (a grid for multiple linked attributes), not the underlying approach.
•Mains-level puzzles (5-7 variables) often require elimination-across-the-full-clue-set when direct placement stalls: list all remaining possibilities for the stuck variable, check each against every other clue, eliminate contradictions.
•Mains puzzles genuinely need more time per puzzle than Prelims (5-8 min vs. 2-3 min) — forcing a Prelims time budget onto Mains difficulty produces rushed errors or incomplete guesses.
Strategy at Mains level
•Selective attempting (solving fewer puzzles with high accuracy) often beats attempting everything with rushed, lower accuracy — the skill is quickly assessing (within ~60-90 seconds) whether a puzzle looks tractable within a reasonable time budget, and strategically deferring intractable-looking ones.
•Computer Aptitude is a genuinely different skill category — direct factual/conceptual recall (number systems, computer terminology, basic networking), not pattern-recognition-based like puzzles. Needs dedicated content study, not repetition-based puzzle practice.
Building durable skill
•The goal of extensive practice isn't memorizing puzzle-type solutions to pattern-match — it's internalizing the underlying method deeply enough to apply reliably to a genuinely novel puzzle structure never seen before, which well-designed exams specifically test for.
•An error log only changes future performance if reviewed AS A SET looking for recurring patterns across entries — logging each mistake independently without cross-referencing misses the same specific, fixable habit (e.g., a consistent off-by-one counting error) recurring across multiple attempts.