How Coding Questions Are Set

A coding question shows you how one or two words are written in a secret code, and then asks you to write a new word in the same code. The setter has used one fixed rule on every word. Find that rule from the examples and use it once more. That is the whole job.

The usual wording looks like this:

"In a certain code language, 'CASH' is written as 'ECUJ'. How will … be written in that language?"

"Is written as" and "is coded as" mean the same thing. Sometimes the question runs the other way and gives you a code, asking which word it stands for. That is decoding, and it gets its own part later in this section.

Two kinds of letter code

In this section the code is made of letters, and the rule is one of two kinds.

  • Letter shift. Each letter is swapped for another letter some places away in the alphabet. The code usually has different letters from the word.
  • Rearrangement. No letter is changed. The letters only swap places, so the code has exactly the same letters as the word.

So the very first check takes two seconds. Compare the letters of the word with the letters of its code. Same letters in a new order means a rearrangement. Different letters means a shift.

The first move: word over code

Write the word on your rough sheet. Under it write the code, letter under letter. Under each pair write the move, which is how many steps forward (+) or back (−) that letter went.

For example, 'CASH' is written as 'ECUJ':

| Word | C | A | S | H | |---|:-:|:-:|:-:|:-:| | Code | E | C | U | J | | Move | +2 | +2 | +2 | +2 |

C to E is two steps forward. A to C is two steps, and so are S to U and H to J. Every move is +2, so the rule is "move each letter two steps forward". Any new word now takes a few seconds.

The third row is where the rule shows itself. Read it like this:

  1. All the moves are the same: every letter moves by one fixed amount.
  2. The moves change in a steady way, like +1, +2, +3, or two moves take turns: a growing or a turn-taking shift.
  3. The moves look like nonsense: test opposite letters, or check whether vowels and consonants move differently.

Each of these has its own part below.

Always use the second example

When a question gives two examples, find the rule on the first one and test it on the second before you look at the options. A rule that fits one example and fails the other is not the rule. Setters give two examples so that only one rule fits both, and they often put the answer of the "almost" rule among the options.

Letter Positions, Opposite Letters and Moving Past Z

Every shift question comes down to counting along the alphabet. You need the place of each letter at once, without saying the alphabet under your breath.

Positions A1 to Z26

| A | B | C | D | E | F | G | H | I | J | K | L | M | |:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:| | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 |

| N | O | P | Q | R | S | T | U | V | W | X | Y | Z | |:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:| | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 |

EJOTY

Five letters make good landmarks: E = 5, J = 10, O = 15, T = 20, Y = 25. Every other letter is one or two steps from one of them. S is one before T, so S = 19. Q is two after O, so Q = 17. X is one before Y, so X = 24.

Reverse positions and opposite letters

Counted from the Z end, Z is 1st, Y is 2nd and A is 26th. A letter's reverse position is 27 minus its position. G is 7th from the front, so it is 20th from the back.

Fold the alphabet in half and each letter lands on its opposite:

| A | B | C | D | E | F | G | H | I | J | K | L | M | |:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:| | Z | Y | X | W | V | U | T | S | R | Q | P | O | N |

A letter and its opposite always add up to 27. If you forget a pair, take the position from 27. The opposite of R(18) is the 9th letter, I.

Moving past Z

The alphabet runs round in a circle, so after Z comes A again. When a move takes you past 26, take away 26.

For example: X + 5. X is 24, and 24 + 5 = 29. Then 29 − 26 = 3, which is C. On your fingers: Y, Z, A, B, C.

Going back past A works the other way round. If you get 0 or a minus number, add 26. C − 4 gives 3 − 4 = −1, and −1 + 26 = 25, which is Y.

You get Do this Letter
27 27 − 26 = 1 A
30 30 − 26 = 4 D
0 0 + 26 = 26 Z
−2 −2 + 26 = 24 X

A big move is a small move the other way

From B to Y is +23 if you count forward, and −3 if you count back. Both land on Y, because 23 + 3 = 26. Setters nearly always mean the small one. So when a move comes out bigger than 13, turn it round: +23 is really −3, and +20 is really −6.

[SSC Tip] Write A1 to Z26 across the top of your rough sheet when the Reasoning section opens. It takes half a minute, and then every coding question is reading, not counting.

Solved Questions

Question 1

In a certain code language, 'CORN' is written as 'EQTP' and 'BEAN' is written as 'DGCP'. How will 'AUNT' be written in that language?

(a) CWPV (b) CWOV (c) BVOU (d) CXPV

Answer. Put CORN over EQTP. C → E is +2, O → Q is +2, R → T is +2, N → P is +2. Now check BEAN: B → D, E → G, A → C, N → P. All +2 again, so the rule is safe. Move every letter of AUNT two steps forward: A → C, U → W, N → P, T → V. So the answer is (a) CWPV.

Time check: about 15 seconds once the moves are written down.

Question 2

In a certain code language, 'BUSY' is coded as 'ATRX' and 'CHIP' is coded as 'BGHO'. How will 'BEND' be coded in that language?

(a) CFOE (b) ADMC (c) ADNC (d) ZDMC

Answer. Here every code letter comes just before its word letter in the alphabet. B → A, U → T, S → R, Y → X. Each one goes back a step, so the move is −1. CHIP → BGHO is −1 for every letter too. For BEND: B → A, E → D, N → M, D → C. CFOE is what you get if you move forward by mistake. So the answer is (b) ADMC.

Question 3

In a certain code language, 'ARMY' is written as 'EVQC' and 'CLAW' is written as 'GPEA'. How will 'AWAY' be written in that language?

(a) EZEB (b) EAED (c) DZDB (d) EAEC

Answer. A → E is +4 and R → V is +4. Y is the tricky one: Y(25) + 4 = 29, and 29 − 26 = 3, which is C. CLAW agrees, with W(23) + 4 = 27, and 27 − 26 = 1, which is A. Now AWAY: A → E, W → A, A → E, Y → C. Count the W once on your fingers if you like: X, Y, Z, A is four steps. EZEB is the slip of stopping one short every time the count crosses Z. So the answer is (d) EAEC.

Five Kinds of Letter Shift

Once the moves are written under the letters, nearly every shift code falls into one of five kinds.

1. The same shift for every letter

'BAIT' is written as 'CBJU'. B → C, A → B, I → J, T → U. Every move is +1. The fixed shift can also go backward, such as −2 for every letter, and it may wrap past Z or past A.

2. A growing shift

'BEST' is written as 'CGVX'. B → C (+1), E → G (+2), S → V (+3), T → X (+4). The move grows by one at each letter. The same idea comes in other forms: +2, +4, +6 … or +1, +3, +5 …, or a shift that runs backward, −1, −2, −3 … Treat the row of moves as a small number series and carry it on.

3. Two shifts taking turns

'COLT' is written as 'DNMS'. C → D (+1), O → N (−1), L → M (+1), T → S (−1). The 1st and 3rd letters move one way and the 2nd and 4th move the other way. Here the move depends on the place, not on the letter. In a five-letter word the first move is used at places 1, 3 and 5.

4. Opposite letters

'AGE' is written as 'ZTV'. The moves come out as +25, +13 and +17, which make no sense. Add each letter to its code letter instead: A(1) + Z(26) = 27, G(7) + T(20) = 27, E(5) + V(22) = 27. Each letter has been swapped for its opposite. Whenever the row of moves looks random, try this test first.

5. Vowels and consonants moved differently

'CANOE' is written as 'BCMQG'. C → B (−1), A → C (+2), N → M (−1), O → Q (+2), E → G (+2). By place the moves are −1, +2, −1, +2, +2, which is no pattern at all. By letter it is clear: the vowels A, O and E all move +2, and the consonants C and N move −1.

Telling them apart

The row of moves The kind
all the same 1. same shift
changes steadily 2. growing shift
two values, by odd and even places 3. taking turns
looks random, but letter + code letter = 27 4. opposite letters
one value for vowels, another for consonants 5. vowels and consonants

Kinds 3 and 5 can look alike. In a word such as 'BOXER', consonants and vowels already take turns, so "+1, −1 by place" and "consonants +1, vowels −1" give the same code. Only a second word that breaks that beat can tell them apart: one with two vowels or two consonants side by side, or one that starts with a vowel. The traps part at the end of this section has a question built on exactly this.

Solved Questions: Growing Shifts, Turns and Opposites

Question 4

In a certain code language, 'CHART' is coded as 'DJDVY' and 'BLEND' is coded as 'CNHRI'. How will 'ANGLE' be coded in that language?

(a) BOHMF (b) BPJOJ (c) BPJPJ (d) BPJPK

Answer. Write the moves under CHART: C → D (+1), H → J (+2), A → D (+3), R → V (+4), T → Y (+5). The move grows by one at each letter. BLEND gives the same row: B → C, L → N, E → H, N → R, D → I. Now ANGLE: A + 1 = B, N + 2 = P, G + 3 = J, L + 4 = P, E + 5 = J. BOHMF moves every letter by +1 and misses the growing moves. So the answer is (c) BPJPJ.

Question 5

In a certain code language, 'BLAST' is written as 'DKCRV' and 'COUNT' is written as 'ENWMV'. How will 'CLOTH' be written in that language?

(a) EKQSJ (b) ENQVJ (c) EKQUJ (d) DKPSI

Answer. B → D is +2, L → K is −1, A → C is +2, S → R is −1, T → V is +2. Two moves take turns: +2 at the 1st, 3rd and 5th places and −1 at the 2nd and 4th. B and L are both consonants and still move differently, so this goes by place, not by letter. COUNT agrees: C → E, O → N, U → W, N → M, T → V. For CLOTH: C → E (+2), L → K (−1), O → Q (+2), T → S (−1), H → J (+2). So the answer is (a) EKQSJ.

Question 6

In a certain code language, 'BALD' is coded as 'YZOW' and 'CRIB' is coded as 'XIRY'. How will 'ACORN' be coded in that language?

(a) YWKHL (b) ZXLIN (c) ZXLIM (d) ZXKIM

Answer. The moves look wild: B → Y is +23 and L → O is +3. When the moves jump about like this, add each letter to its code letter. B(2) + Y(25) = 27, A(1) + Z(26) = 27, L(12) + O(15) = 27, D(4) + W(23) = 27. Each letter is swapped for its opposite. CRIB → XIRY checks out the same way, for example R(18) + I(9) = 27. For ACORN: A ↔ Z, C ↔ X, O ↔ L, R ↔ I, N ↔ M. So the answer is (c) ZXLIM.

Solved Questions: Vowels, Consonants and Backward Runs

Question 7

In a certain code language, 'BEAD' is written as 'AFBC' and 'CUBE' is written as 'BVAF'. How will 'ACTOR' be written in that language?

(a) BDUPS (b) BBSPQ (c) ZDUNS (d) BBSQQ

Answer. BEAD over AFBC: B → A (−1), E → F (+1), A → B (+1), D → C (−1). By place that is −1, +1, +1, −1, which is no pattern. Look at the letters instead. E and A are vowels and both move +1. B and D are consonants and both move −1. CUBE agrees: U → V and E → F go up, C → B and B → A go down. In ACTOR the vowels A and O go up one (A → B, O → P) and the consonants C, T and R go down one (C → B, T → S, R → Q). So the answer is (b) BBSPQ.

Question 8

In a certain code language, 'COAT' is coded as 'BMXP' and 'BARN' is coded as 'AYOJ'. How will 'AXIS' be coded in that language?

(a) BZLW (b) ZVFP (c) ZWFO (d) ZVFO

Answer. C → B is −1, O → M is −2, A → X is −3 (A goes back past the start: Z, Y, X), and T → P is −4. The letters go back by 1, 2, 3, 4. BARN agrees: B → A, A → Y, R → O, N → J. For AXIS: A − 1 wraps round to Z, X − 2 = V, I − 3 = F, S − 4 = O. BZLW comes from moving forward instead of back. So the answer is (d) ZVFO.

Rearrangement Codes

In a rearrangement code no letter is changed. The letters of the word are only put in a new order. So once you see that the code has exactly the letters of the word, stop counting steps and look at where each letter went.

The common patterns

Here is one word, 'CRUISE', put through each pattern you are likely to meet.

Pattern How it works 'CRUISE' becomes
Reversed read the word from the end ESIURC
Halves reversed cut in the middle and reverse each half: CRU, ISE → URC, ESI URCESI
Pairs swapped swap the 1st and 2nd, the 3rd and 4th, and so on: CR, UI, SE → RC, IU, ES RCIUES
First and last swapped the middle stays put ERUISC
All but the last reversed reverse CRUIS, keep E at the end SIURCE
Alphabetical order sort the letters from A to Z CEIRSU

Words with an odd number of letters

Halves and pairs need care when the length is odd. When a seven-letter word is cut into halves, the middle (4th) letter usually stays where it is and the three letters on each side are reversed. When pairs are swapped in a five-letter word, the last letter has no partner and stays at the end. Check your example to see how the setter has handled it.

How to spot the pattern

  1. Write the word and number its letters 1, 2, 3 …
  2. Under each code letter, write the number that letter had in the word. For 'CRUISE' → 'URCESI' you get 3 2 1 6 5 4.
  3. Read the numbers. 6 5 4 3 2 1 is a full reverse. 3 2 1 6 5 4 is halves reversed. 2 1 4 3 6 5 is pairs swapped.

The number pattern does not care which letters were used, so it carries straight over to any word of the same length. Use this whenever the pattern does not jump out at you.

Test on both examples

One example can sometimes fit two patterns, especially a short word or one whose letters are nearly in alphabetical order already. Test your pattern on the second example before you choose. Alphabetical order is the odd one out here: it works by the letters, not by their places, so check it whenever the code looks "tidy".

Solved Questions: Rearrangements

Question 9

In a certain code language, 'CLAMP' is written as 'PMALC' and 'BUNCH' is written as 'HCNUB'. How will 'APRON' be written in that language?

(a) NOPRA (b) ONRPA (c) NORPA (d) NORAP

Answer. The codes use exactly the letters of the words, so nothing has been shifted. The letters have only moved. Read CLAMP from the back: P, M, A, L, C. That is PMALC. BUNCH from the back is HCNUB. APRON read from the back is N, O, R, P, A. So the answer is (c) NORPA.

Question 10

In a certain code language, 'BEACON' is coded as 'AEBNOC' and 'CASTLE' is coded as 'SACELT'. How will 'BRANCH' be coded in that language?

(a) ARBHCN (b) HCNARB (c) NCHBRA (d) ARBNCH

Answer. BEACON has six letters, so cut it in the middle: BEA | CON. BEA reversed is AEB and CON reversed is NOC, and side by side they make AEBNOC. CASTLE → CAS | TLE → SAC | ELT agrees. Now BRANCH: BRA | NCH, which turns into ARB | HCN. HCNARB would be the whole word reversed, and the examples do not do that. So the answer is (a) ARBHCN.

Question 11

In a certain code language, 'CLIENT' is written as 'LCEITN' and 'BUCKET' is written as 'UBKCTE'. How will 'ACTION' be written in that language?

(a) CATINO (b) NOITCA (c) CAITON (d) CAITNO

Answer. Pair up the letters of CLIENT: CL, IE, NT. Each pair is turned round: LC, EI, TN. BUCKET → UB, KC, TE does the same. ACTION pairs up as AC, TI, ON, which turn into CA, IT, NO. So the answer is (d) CAITNO.

Time check: about 15 seconds once you see the pairs.

Question 12

In a certain code language, 'CHARM' is coded as 'MHARC' and 'BRUSH' is coded as 'HRUSB'. How will 'ALBUM' be coded in that language?

(a) MUBLA (b) MLBUA (c) LABUM (d) ALBMU

Answer. Compare letter by letter. In CHARM → MHARC the middle part HAR has not moved at all. Only C and M have changed places. BRUSH → HRUSB is the same: B and H swap and RUS stays. So in ALBUM, A and M swap and LBU stays where it is. MUBLA is ALBUM reversed, but the examples keep the middle as it is. So the answer is (b) MLBUA.

Question 13

In a certain code language, 'CANDY' is written as 'DNACY' and 'BOARD' is written as 'RAOBD'. How will 'BLADE' be written in that language?

(a) EDALB (b) DABLE (c) DALBE (d) ADLBE

Answer. In both codes the last letter is still last: Y in DNACY and D in RAOBD. The first four letters are reversed: CAND → DNAC and BOAR → RAOB. For BLADE, keep E at the end and reverse BLAD to get DALB. EDALB reverses all five letters and moves the E. So the answer is (c) DALBE.

Question 14

In a certain code language, 'CRISP' is coded as 'CIPRS' and 'BEACH' is coded as 'ABCEH'. How will 'CLAIM' be coded in that language?

(a) ACILM (b) ACLIM (c) CAMLI (d) MLICA

Answer. CRISP → CIPRS can be read in two ways. It could be the 1st, 3rd and 5th letters (C, I, P) followed by the 2nd and 4th (R, S). Or it could be the letters in alphabetical order: C, I, P, R, S. BEACH settles it. Its code ABCEH is in alphabetical order, while the odd-then-even reading would give BAHEC. So the rule is alphabetical order. CLAIM sorted is A, C, I, L, M. CAMLI is the trap left by the odd-then-even reading. So the answer is (a) ACILM.

Decoding: From the Code Back to the Word

Sometimes you are given a code and asked for the word. For example:

"In a certain code language, 'BRAWN' is written as 'DTCYP'. Which word is written as 'EQTG' in that language?"

Find the rule from the example in the usual way. BRAWN → DTCYP is +2 for every letter. Now run it backwards. To undo a move, make the opposite move.

| Code | E | Q | T | G | |---|:-:|:-:|:-:|:-:| | Undo | −2 | −2 | −2 | −2 | | Word | C | O | R | E |

The word is CORE. To check, code CORE with +2: you get EQTG again.

Undoing each kind of shift

  • Same shift: move every letter the same amount the other way.
  • Growing shift: undo each place on its own. If the code used +1, +2, +3, +4, take away 1, 2, 3 and 4.
  • Two shifts taking turns: undo each one at its own places.
  • Opposite letters: the opposite of an opposite is the letter you started with, so decoding is the same as coding.
  • Vowels and consonants: be careful. A code letter could have come from a vowel or from a consonant. Try both ways back, keep the one where the letter type matches, and let the options decide if both work.

Undoing a rearrangement

Most rearrangements undo themselves. Reverse a word twice and you are back where you began. The same is true for halves reversed, pairs swapped, first and last swapped, and all but the last reversed. So to decode, apply the very same move to the code. 'CRUISE' reversed is 'ESIURC', and 'ESIURC' reversed is 'CRUISE' again.

Alphabetical order is different. Once the letters are sorted, nothing tells you their old order, so it does not suit a decoding question.

Use the options, but finish the word

In decoding the options are usually words or near-words. After two letters most of them drop out. Still work out every letter, because setters like options that differ only near the end. The most common wrong option is the one you get by applying the rule forwards again instead of backwards, so expect to see it.

Solved Questions: Decoding

Question 15

In a certain code language, 'CARGO' is written as 'FDUJR'. Which word is written as 'FDQDO' in that language?

(a) IGTGR (b) CANAL (c) DBOBM (d) CAMAL

Answer. CARGO → FDUJR: C → F, A → D, R → U, G → J, O → R. Every letter moves +3. Here we have the code and need the word, so undo the move and take every letter three steps back. F → C, D → A, Q → N, D → A, O → L. That spells CANAL. IGTGR is what you get by moving forward again, which codes the code a second time. So the answer is (b) CANAL.

Question 16

In a certain code language, 'BISHOP' is written as 'SIBPOH'. Which word is written as 'UOBECN' in that language?

(a) NCEBOU (b) ECNUOB (c) BOUECN (d) BOUNCE

Answer. BISHOP → SIBPOH: the halves BIS and HOP are each reversed, giving SIB and POH. Reversing a half twice puts it back, so to decode we do the same move again. UOBECN splits into UOB and ECN. Reverse each: BOU and NCE. BOUECN undoes only the first half. So the answer is (d) BOUNCE.

Traps in Letter Codes

Trap 1: a rule that fits only one example

'COG' is written as 'DQJ'. The moves are +1, +2, +3. That could be a growing shift (+1, +2, +3, +4, +5 …) or a set of three moves that repeats (+1, +2, +3, +1, +2 …). For a three-letter word the two readings agree. For a five-letter word they don't: the 4th and 5th letters get +4 and +5 under one reading, and +1 and +2 under the other. This is why a careful setter adds a second, longer example. Find the rule on one example, then prove it on the other.

The same thing happens with "taking turns" against "vowels and consonants", and with rearrangements on short words. Whenever two rules fit the first example, the second example is there to choose between them.

Trap 2: counting the shift the wrong way

'CAP' is written as 'ZXM'. Counting forward from C to Z gives +23. Counting from the code back to the word, Z to C, gives +3. The real move is −3: C goes back three steps (B, A, Z), and so do A (Z, Y, X) and P (O, N, M). Always count from the word to the code, and if the move is bigger than 13, turn it round.

Trap 3: the wrap

X + 4 is B, not A. Count it: X → Y → Z → A → B. Stopping one short is the usual slip when a count crosses from Z to A. Doing the sum is safer: 24 + 4 = 28, and 28 − 26 = 2, which is B.

Trap 4: places in taking-turns codes

When two moves take turns, a letter's code depends on its place. A word with a double letter, like the two S's in 'BLESS', can give two different code letters. Number the places before you start.

Trap 5: options that are one letter off

Wrong options usually match the right answer in every letter but one. Work out the whole code, not just the first two letters, and check the last letter as carefully as the first.

[SSC Tip] The Reasoning section has 25 questions on its own locked 15-minute clock, with +2 for a right answer and −0.5 for a wrong one. A letter code that you have checked on both examples is one of the safest +2s in the section, so do the check.

Solved Questions: Common Mistakes

Question 17

In a certain code language, 'COMET' is coded as 'DNNDU' and 'BEAST' is coded as 'CDZTU'. How will 'ANKLE' be coded in that language?

(a) BMLKF (b) ZOLMF (c) ZOLMD (d) BOLMD

Answer. Start with COMET → DNNDU: C → D (+1), O → N (−1), M → N (+1), E → D (−1), T → U (+1). It looks like +1 and −1 taking turns. But in COMET the consonants and vowels also take turns, so a second rule fits just as well: consonants +1, vowels −1. BEAST decides between them. B → C (+1), E → D (−1), A → Z (−1), S → T (+1), T → U (+1). A sits at the 3rd place, where taking turns would need +1, but it goes back. So taking turns is out, and consonants +1, vowels −1 fits every letter. For ANKLE: A → Z, N → O, K → L, L → M, E → D. BMLKF is the trap answer from COMET alone. So the answer is (c) ZOLMD.

Question 18

In a certain code language, 'CROWN' is written as 'ZOLTK' and 'BADGE' is written as 'YXADB'. How will 'CHESS' be written in that language?

(a) ZEBPP (b) FKHVV (c) ZEBQQ (d) AFCQQ

Answer. C → Z looks like a big jump of +23. A move bigger than 13 is really a small move the other way: C goes back three steps, to B, A, Z. R → O and O → L are −3 as well, so the rule is −3, not +23. BADGE agrees: B → Y, A → X, D → A, G → D, E → B. Now CHESS: C → Z, H → E, E → B, S → P, S → P. FKHVV comes from counting +3 from the code to the word, which is the right size in the wrong direction. So the answer is (a) ZEBPP.

Time check: under 25 seconds, most of it spent turning +23 into −3.