Strategy
Score 140+ in NDA English with ~48 hours of prep
50 English questions × 4 marks − 1.33 per wrong. Here's the math, then the three buckets.
- marks out of 200
- 140+
- attempts of 50
- 40
- accuracy needed
- 85%
- total prep time
- ~48 h
The arithmetic of 140+
NDA GAT’s English half has 50 questions worth 4 marks each, with −1.33 per wrong. To net 140 marks you need:
| Accuracy | Attempts | Correct | Wrong | Net marks | Result |
|---|---|---|---|---|---|
| 75% | 40 | 30 | 10 | ~107 | Miss |
| 85% | 40 | 34 | 6 | ~128 | Close ✓ |
| 90% | 40 | 36 | 4 | ~139 | Target ✓ |
Target: attempt 40 questions at ≥85% accuracy. Skip the 10 you’re unsure of. The arithmetic for English is more forgiving than Maths’ — 4 marks per correct vs 2.5, but a steeper 1.33-penalty per wrong, so accuracy still matters more than coverage.
Recall — Vocabulary and Idioms (402 q / 45% of bank)
Pure recall — either you know the word/idiom or you don't. No partial credit, no reasoning route. The biggest bucket in the bank, the lowest-difficulty (Vocab 2% HARD, Idioms 3% HARD), and the most leverage per hour of focused word-list work.
The approach
- Build a tested-word list from the live bank (270 underlined words across Synonyms + Antonyms). Read /guide/nda-english/vocab-families for the thematic clusters.
- Drill in 20-word batches with spaced repetition (Anki, Quizlet). Test cold the next day. Words don't repeat across years (only 5 words tested twice in 10 yrs) — so memorise the *families*, not the stems.
- Idioms: 85 unique in 10 yrs. Group by theme (animal, body-part, colour, weather). The literal-meaning trap (Section 4 on /guide/nda-english/traps) is the only consistent distractor.
Vocabulary
316 questions · 2% hard
316 q in 10 yrs. 49% of attempts are EASY. Build a 300-word list across the 13 vocab families. Confusable pairs + Word Definition are too sparse to dedicate time to — they'll fall out of the same word work.
Skip
- Confusable Word Pairs
- Word Definition
Idioms and Phrases
86 questions · 3% hard
86 q / 85 unique idioms. Read them grouped by theme (body-parts, animals, colours, weather, money). The literal interpretation is almost always a wrong option — that's the trap shape NDA reuses.
Drill
- Idiom MeaningDrill
Rule — Grammar and Spotting Errors (223 q / 25%)
Both chapters test the same ~10 grammar rules in different formats. Errors uses underline-the-mistake; Grammar uses fill-in-the-blank, identify-the-correct-sentence, or pick-the-right-connector. The rules are stable — they don't change year to year, so this bucket has the longest shelf life.
The approach
- Master the 5 most-tested rule clusters: subject-verb agreement, tense/verb-form, articles+determiners, prepositions, word choice (affect/effect, fewer/less, who/whom).
- Practice both formats in the same session — a S-V rule answered under Errors format trains the same muscle as the Grammar format. The cross-chapter overlap (27 q for S-V alone) is the highest in English.
- Grammar EXPLODED from 0 q in 2017–2018 to 30+ q/year after 2024. If you only practised pre-2024 papers, you're undertrained on Sentence Completion (30 q) and Connectors (20 q) which are post-2024 inventions.
Spotting Errors
115 questions · 3% hard
115 q. The underline-format that tests rule recognition. Drill the 4 named subtopics; Mixed + No-Error are calibration practice once you're solid on the rest.
Drill
Skip
- Mixed Error Detection
- No Error (Correct Sentence)
Grammar
108 questions · 4% hard
108 q, almost all post-2024. Sentence Completion + Connectors carry half the chapter (50 q). Drill Connectors aggressively — they test the same logic-of-discourse skill that Cloze and PQRS reward.
Skip
- Active and Passive Voice
Reason — RC, Rearrangement, FIB, Cloze (275 q / 31%)
Context-driven. No vocabulary deficit and no rule lookup will save you here — these reward reading a passage carefully and following the logic of discourse. Sentence Rearrangement is the only chapter in NDA English with real HARD load (22% overall, Paragraph Sequencing at 36% HARD).
The approach
- Read 1 RC passage + 1 Cloze + 1 PQRS daily. Time-box: 8 min for RC, 4 min for Cloze, 2 min per PQRS. Don't read for comprehension; read for structure — what's the author claiming, what's the counter, what's the example?
- Paragraph Sequencing (S1–S6) is the hardest single subtopic in NDA English. The skill: locate the *opener cue* (a noun without a referring pronoun, a topic-setting statement) and the *closer cue* (a conclusion or a generalisation). Middle sentences fall into place by transition words (However, Moreover, So, Thus).
- FIB and Cloze look similar but test different things — FIB is sentence-scoped (vocabulary + collocation), Cloze is passage-scoped (logic + transition). Don't conflate them.
Sentence Rearrangement
114 questions · 22% hard
114 q. PQRS is the bread (92 q), S1–S6 is the butter (22 q at 36% HARD). The hardest single subtopic in NDA English. Train opener/closer cue recognition.
Reading Comprehension
61 questions · 7% hard
61 q, set-bound (passages have 4–8 q each — answer them in order, not skipping). Inferential (43 q) dominates and is harder; Literal (14 q) is the easy plumbing question.
Fill in the Blanks
55 questions · 2% hard
55 q. Sentence-scoped — vocabulary + collocation. Phrasal-verb subset (10 q) is rare; drill the contextual main bucket.
Drill
- Contextual Fill-in-BlankDrill
Cloze Test
45 questions · 0% hard
45 q, zero HARD. Pure transition + discourse-connector reasoning. Often appears as one 5-blank passage. Easiest 5 marks if you can read for structure.
Drill
- Word Selection in PassageDrill
Test-day attempt order
Bank Recall marks first (fast, high-confidence), then Rule, then Reason (slowest, context-heavy). Don’t start with passages — they bleed time.
- 15min
Sweep Recall (Vocab + Idioms)
Scan the paper, attack every Synonym/Antonym/Idiom question first. ~15 q × ~45 sec/q. Target: 12–13 correct in 15 minutes — that's already 50 marks banked. If you don't recognise the word, skip — don't guess; the −1.33 penalty kills.
- 20min
Sweep Rule (Errors + Grammar)
Attempt every Spotting Errors and Grammar question. ~12 q × ~90 sec/q. Pre-2024 you'd see 5 Grammar; in 2025+ you'll see ~15. Target: 10–11 correct.
- 25min
Reason (RC + Cloze + PQRS + FIB)
Tackle passages last — they're time-intensive. RC first (4–8 q per passage answers in batch), then Cloze (5-blank passage), then PQRS one-by-one. S1–S6 paragraph sequencing only if you have time spare — it's the bank's only chapter at 36% HARD.
Never guess on Vocab or Idioms. The 1.33 penalty wipes out 33% of a correct answer. If you don’t know the word, skip — don’t pick.
Time investment plan
| Bucket | Hours | Outcome |
|---|---|---|
| Recall — Vocab + Idioms | 16 | ~16 marks/paper |
| Rule — Errors + Grammar | 10 | ~11 marks/paper |
| Reason — Rearr + RC + FIB + Cloze | 14 | ~13 marks/paper |
| Past papers, timed | 8 | Calibration + speed |
| Total | 48 | Target: 130–140 marks |
That’s about 7 weeks at 7 hours/week. Recall takes the most hours because vocabulary breadth doesn’t come fast — but it’s also the most reliable bucket for marks-per-hour.
Start with Vocabulary — the biggest bucket
316 questions in the bank, 12+ marks per paper, 12 hours of work.