Numeric and Letter Series

Overview

Numeric and Letter Series appears about 3.5 times per paper across the last four AFCAT solved papers, placing it in the highest weight band of Reasoning and Aptitude.

Numeric and Letter Series is the single largest topic in the AFCAT Reasoning and Military Aptitude block. Across the four solved papers from 2022 to 2025, an average of 3.5 questions per paper come from this topic — between ten and eleven marks at three marks each, with a one-mark penalty for a wrong tick. That alone makes it the highest-yield Reasoning topic to drill before the exam.

The reason this topic is dense is also why it is winnable. Series questions reward a candidate who has internalised a small catalogue of patterns — perhaps fifteen to twenty rule shapes — and who can run a fixed diagnostic checklist on any new stem in under a minute. AFCAT does not invent new mathematics for series; it recycles the same rule families year after year, dressing them in different starting numbers and different missing positions. A candidate who has memorised the squares to thirty, the cubes to fifteen, the first fifteen Fibonacci terms and the primes to one hundred has already solved half the bank by recognition alone.

The other half rewards process. AFCAT's favourite trap is the planted pattern: a stem where +5, +6 fits the first jump, where ×2 fits the next, and where the actual rule is neither — it is ×n+k with shrinking n, or differences-of-differences, or two interleaved sub-series. This page lays out the recognition catalogue, the four-step method that catches the planted traps, ready-recall tables, fifteen worked examples covering every major pattern family, and the time budget that lets you bank all eleven marks inside the ninety-second slot AFCAT effectively gives you per Reasoning item.

Why series is the largest Reasoning topic in AFCAT

The AFCAT Reasoning and Military Aptitude block carries between twenty-five and thirty questions per paper, contributing about twenty-seven per cent of the total paper. Of that block, series alone accounts for around fourteen per cent — the largest slice. Coding-decoding comes second at twelve per cent, analogies at ten, direction sense at nine and Venn diagrams at eight.

Two factors drive the dominance of series. First, series is the cheapest item type for the paper-setter to generate: every fresh combination of starting term and rule produces a new question without any need for fresh prose, a fresh figure or fresh culture-loaded vocabulary. Second, series scales cleanly in difficulty by simply changing the rule family — from a transparent arithmetic progression at the easy end to a shrinking multiplier with an added constant at the hard end — without any change in stem length. That makes the topic the natural workhorse for filling the Reasoning section.

The takeaway is operational. Build the recognition library first. Build the four-step process second. Then drill enough mixed sets that the four-step process runs without conscious effort, leaving working memory free to handle the one trap question per paper that does not respond to standard rules.

The four-step method

Every series problem should run through the same checklist in the same order. The order matters: it puts the cheapest test first and the most expensive test last, so most items resolve in step one or two.

1. Step 1 — Compute first differences. Write the gaps between consecutive terms beneath the stem. If they are constant, the rule is arithmetic. If they form a recognisable progression of their own (squares, cubes, doubling, primes), the rule is a differences-driven one and the answer falls out immediately.
2. Step 2 — Test ratios. If the differences are not clean, divide each term by the previous one. A constant ratio is a geometric progression. A ratio that itself decreases — say 5, then 4, then 3 — signals the ×n+k family where n shrinks.
3. Step 3 — Check for two interleaved sub-series. Mark the odd-position terms with circles and the even-position terms with squares. Apply step one to each sub-series independently. AFCAT plants interleaved series often enough that this should be a reflex on any stem of six terms or more.
4. Step 4 — Match against the standard pattern table. If the first three steps have not cracked it, look for the stem to resemble a square, cube, factorial, Fibonacci or prime sequence, or a multiply-and-add scheme. Inspection at this stage is fast because the recognition library has done the heavy lifting.

Standard pattern catalogue

This is the recognition library. Memorise the example next to each rule; the example is the visual signature you scan for when reading a fresh stem.

Run the catalogue from top to bottom on any stem you cannot crack with the four-step method. One of these signatures almost always fits.

Number series — non-standard step patterns

AFCAT's most consistent trap is the multiply-and-add family with a shrinking multiplier. The 2024 paper used 1, 6, 25, 76, ?, 154 — a stem that disguises a ×5+1, ×4+1, ×3+1, ×2+1, ×1+1 rule behind innocuous first terms. The detection signal is always the same: compute the ratios of consecutive terms and watch for a monotonic shrink.

For any stem where the first differences are not clean and the ratios are not constant, the next test is whether the ratios themselves form a clean sequence — most of AFCAT's harder series fall under one of the rows above.

Letter series — convert to numbers first

The base technique is to map A = 1, B = 2 and so on through Z = 26, write the position numbers under the stem, and apply the number-series method. Reverse-mapping at the end converts the answer back to a letter. AFCAT's letter-series questions fall into five repeating shapes:

• Constant gap. A, C, E, G (gaps of 2 in position); C, F, I, L (gaps of 3); A, D, G, J (gaps of 3). The gap is the constant added each step.
• Increasing gap. A, B, D, G, K (gaps of 1, 2, 3, 4 — gap itself in arithmetic progression). After K (position 11), the next gap is 5, giving P (position 16).
• Mirror-pair. A, Z; B, Y; C, X; D, W — the two positions sum to 27. Used in stems like AZ, BY, CX, ?, where the answer is DW.
• Group-of-three or group-of-four. ABCD, EFGH, IJKL, MNOP — clusters of four with a constant shift of 4 between cluster starts.
• Alternating two letter-series. A, M, B, N, C, O — odd positions advance by 1, even positions advance by 1 starting from M.

The mirror column is built by subtracting the position from 27. Memorising the first column to the point of instant recall is the single highest-yield rote investment for letter-series items.

Mixed alphanumeric series

AFCAT samples alphanumeric stems lightly — perhaps one paper in four carries one. The pattern is a letter paired with a number in each term, with the letter following one rule and the number following another. Treat the two streams as independent.

For a stem like A1, C4, E9, G16, I25 the letters follow A, C, E, G, I — constant gap of two in position — and the numbers follow 1, 4, 9, 16, 25 — squares of n. The next term is K36. The technique is to split, solve each stream with the four-step method, and recombine.

The variant to watch is the symbol-letter-number triplet. A stem such as P3@, R6#, T9$ embeds a third stream of symbols. The symbol stream is rarely a real rule; usually it cycles or repeats. Identify the cycle in the symbol stream first and then solve the letter and number streams in the normal way.

Series with two missing terms

The AFCAT stem 1, 6, 25, 76, ?, 154 carries two unknowns at positions five and six. Two-missing-term stems are common because they give the candidate a built-in cross-check: the proposed rule must hit both blanks consistently with the surrounding given terms.

The technique is to use the given terms on both sides of the unknown to triangulate the rule. In the example, the last given term 154 sits two steps from the last computed term 76. If the rule is ×2+1, ×1+1, then 76 × 2 + 1 = 153 and 153 × 1 + 1 = 154 — both blanks resolve consistently, confirming the rule. If a candidate rule satisfies the first blank but contradicts the last given term, discard the rule and re-test.

For arithmetic and geometric stems, two-missing-term items are even easier because the rule is detectable from any two non-adjacent given terms. If you know the term at position three and the term at position six, the rule lets you compute the average step in between by interpolation, then walk back to fill both blanks.

Wrong-term identification

The wrong-term variant gives a complete series of six or seven terms and asks the candidate to identify the one term that breaks the underlying rule. The technique is to run the standard four-step method on the longest unbroken stretch first, then check each candidate term against the established rule.

Consider the stem 2, 5, 10, 17, 27, 37. First differences are 3, 5, 7, 10, 10. The pattern of odd numbers 3, 5, 7 breaks at 10. The clean rule is n²+1 — 2, 5, 10, 17, 26, 37 — so the planted error is 27 (it should be 26). Spotting the irregularity in the differences is faster than re-deriving the rule from scratch.

A second variant is the all-prime stem with one composite slipped in. 11, 13, 15, 17, 19 — here 15 is the intruder; the true series is 11, 13, 17, 19 (skipping 15) with the planted error at position three. Recognition tables for primes catch these immediately.

Series in figure form

Figure-series items show four or five figures changing across the row, with the candidate asked to pick the figure that continues the pattern. The transformations to watch for are rotation by a constant angle (45, 90 or 180 degrees clockwise or anti-clockwise), addition of one element per step (a dot, a line, a shaded segment), shading rotating around a fixed arrangement of cells, and shape substitution where the outer shape changes through a cycle.

The technique is to describe each figure in words — "square with one dot at top-left," "square with two dots at top-left and top-right" — and treat the verbal description as the series. That converts the visual problem into a regular series problem the four-step method can handle.

Figure series in AFCAT are usually one or two questions per paper and sit inside the wider non-verbal-pattern cluster covered separately. Do not over-invest in figure-series drills at the cost of numeric and letter series, where the marks density is much higher.

Quick-recognition recall tables

Memorise these five tables to the point of instant recall. Each table converts a slow computation into a one-second lookup, which is the difference between converting a series item inside the time budget and burning two minutes on it.

Squares of 1 to 30.

Cubes of 1 to 15.

Primes up to 100.

2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47, 53, 59, 61, 67, 71, 73, 79, 83, 89, 97. That is twenty-five primes. The composites that AFCAT plants as intruders most often are 15, 21, 25, 27, 33, 35, 49, 51, 55, 57, 63, 65, 77, 87, 91, 93, 95 and 99.

First fifteen Fibonacci numbers.

1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, 233, 377, 610. Each term is the sum of the previous two starting from the seed pair 1, 1.

Factorials 1! to 7!.

Factorials grow fast enough that AFCAT rarely uses 8! or higher inside a stem; recognise 5040 and the candidate rule is settled.

Time budget and pacing

AFCAT gives 120 minutes for 100 questions — a notional 72 seconds per item. Reasoning, with its mix of figure-based and verbal items, is the section where pacing matters most. For series specifically, the working budget is sixty to seventy-five seconds per item, structured as follows.

1. Seconds 0–15: Read the stem, copy the numbers (or position values for letters), compute first differences directly below.
2. Seconds 15–30: If the differences are clean, identify the rule, walk it forward to the missing term, tick the answer.
3. Seconds 30–45: If the differences are not clean, compute ratios. If a clean ratio appears, apply it; if ratios shrink monotonically, test the ×n+k family.
4. Seconds 45–60: If neither differences nor ratios resolve it, check for two interleaved sub-series.
5. Seconds 60–75: Match against the recognition catalogue. If still unresolved, mark for review and move on — never burn more than ninety seconds on a single series item.

Over a 25-question Reasoning section with eight items in the series-coding-analogies cluster, this pacing leaves six to seven minutes of buffer for tougher figure-based items later in the block. The cost of overrunning on one series stem is the buffer for two figure items at the back of the section.

AFCAT-specific trap patterns

Three traps repeat across the AFCAT series bank. Knowing them by name is the cheapest defence.

1. The vanishing-multiplier trap. Stems like 1, 6, 25, 76 fit a ×6, then ×4 ratio at the first jump, which looks like a sloppy geometric progression. The actual rule is ×5+1 with a shrinking n. Defence: if your ratio drops by more than 20 per cent between consecutive jumps, the rule is multiply-and-add.
2. The square-with-offset trap. Stems like 2, 5, 10, 17, 26 read like an arithmetic progression with steps 3, 5, 7, 9 — and they are, but the underlying rule is n²+1 not just incrementing odd differences. Both descriptions give the same answer, so this is not a wrong-answer trap. It becomes one in stems like 4, 9, 16, 27, 36 where one term is mis-set and the candidate following "differences are odd numbers" misses the deeper square rule that exposes 27 as the planted error.
3. The interleaved-series trap. Stems of six or more terms with no clean overall rule often resolve into two interleaved sub-series. Defence: as soon as the four-step method's step one and step two fail, run step three before any further effort. The check costs five seconds; missing the interleave costs the whole item.

Beyond these three, AFCAT does not innovate. The bank reuses the same nineteen rule shapes from the pattern catalogue in section three, and the trap inventory is fixed. A candidate who has worked thirty mixed-rule series items in the week before the exam will recognise every stem within fifteen seconds.

Worked AFCAT-style examples

Exam-day strategy

1. Run the four-step method on every stem in fixed order — differences, ratios, interleave, catalogue — without skipping ahead.
2. Memorise the recognition tables for squares to 30, cubes to 15, primes to 100, the first fifteen Fibonacci terms and factorials to 7. These five tables resolve more than half the series bank by lookup.
3. Never lock a rule after only two jumps. AFCAT plants rules that fit the first two terms and break later — confirm against at least three consecutive transitions.
4. For letter series, write positions under the letters and treat as a number-series problem. Reverse-map the answer at the end.
5. When ratios shrink monotonically between consecutive jumps, test the multiply-and-add family with shrinking multiplier before anything else.
6. Stems with six or more terms warrant a five-second interleave check the moment the standard four-step method fails — split into odd and even position sub-series.
7. Cap each item at 75 seconds. If unresolved, eliminate options that fit no recognisable rule, pick the most plausible survivor, and move on. The expected value of a two-option guess at +3 / −1 is positive.
8. On wrong-term identification, compute differences for the full stem before assuming the first three terms are correct. The planted error usually sits at position three or four.
9. On two-missing-term stems, use the last given term as a cross-check on the candidate rule — both blanks must resolve consistently with the given terms on either side.
10. Drill thirty mixed-rule series items in the final week before the exam. Series rewards pattern fluency built by exposure, not slow deliberation under timed conditions.

Frequently asked questions

Related topics