Coding-Decoding is the art of reading a secret language — one where words, letters, or numbers have been transformed according to a hidden rule, and your job is to crack that rule and apply it.
Think of it like a post office cipher from a spy movie. The sender and receiver both know the rule. You, the exam candidate, are the codebreaker who figures out the rule from examples given and then applies it to the unknown.
In the IBPS RRB Office Assistant exam, this topic appears in the Reasoning section and is considered a high-scoring area because the questions are short, rule-based, and once you identify the pattern, the answer follows mechanically. There is no guesswork if you are systematic.
Here is the core idea: every coding problem gives you at least one known input-output pair (the example), from which you reverse-engineer the transformation rule. Then you apply the same rule to the new input to get the coded (or decoded) output.
A classroom analogy: Imagine a teacher who consistently gives every student a nickname by adding one letter to the front of their name. "Rohit" becomes "ARohit", "Priya" becomes "APriya". Once you notice this, you can instantly decode "ASuresh" — it is "Suresh". Coding-Decoding works exactly this way: find the consistent rule, apply it without hesitation.
The transformation can happen to:
The beauty of this topic is that the rule is always consistent. There are no trick exceptions. Once cracked, it holds for every element in that question.
This is the most common type in IBPS RRB. Each letter in the original word is shifted a fixed number of positions forward or backward in the English alphabet.
Forward shift (+n): A→B→C→...→Z→A (wraps around) Backward shift (−n): Z→Y→X→...→A→Z (wraps around)
How to identify the shift: Take the first letter of the given word and the first letter of the coded word. Count the gap.
Example from the PYQs: FLOWER → GMPXFS
Confirmed. Shift = +1. Apply to NATURE: N(+1)=O, A(+1)=B, T(+1)=U, U(+1)=V, R(+1)=S, E(+1)=F → OBUVSF
For −1 shift: MOTHER → LNSGDQ
Confirmed. Shift = −1.
Key tool you must have memorised: The alphabet with positions.
| A=1 | B=2 | C=3 | D=4 | E=5 | F=6 | G=7 | H=8 | I=9 | J=10 | K=11 | L=12 | M=13 | |-----|-----|-----|-----|-----|-----|-----|-----|-----|------|------|------|------| | N=14 | O=15 | P=16 | Q=17 | R=18 | S=19 | T=20 | U=21 | V=22 | W=23 | X=24 | Y=25 | Z=26 |
For +3 shift (Caesar cipher variant): A→D, B→E, C→F ... X→A, Y→B, Z→C. This appeared in the MANGO→PDQJR question (M=13, P=16, gap=3).
This type gives you two or three complete sentences with their codes and asks you to identify one specific word's code.
The technique: Comparison by Elimination
This is pure process of elimination — no guessing, no shortcuts needed. Just be systematic.
Example structure:
Here, coded words replace arithmetic operators (+, −, ×, ÷). After substitution, you apply BODMAS (Brackets, Orders, Division, Multiplication, Addition, Subtraction) strictly.
Critical point: Most candidates who get these wrong do not fail at decoding — they fail at BODMAS. The decoding step takes 5 seconds. The arithmetic step is where marks are lost.
Process:
Example: "8 find 16 for 4 joy 3 new 2" where find=+, for=÷, joy=×, new=−
Decoded: 8 + 16 ÷ 4 × 3 − 2
BODMAS: 16 ÷ 4 = 4 → 4 × 3 = 12 → 8 + 12 = 20 → 20 − 2 = 18
Similar to Type 2 but uses number codes for words. Comparison by elimination is the method. The number codes are arbitrary — do not try to find a pattern in the numbers themselves. Just eliminate.
Some questions code based on letter position in the alphabet. For example, ABCD might become 1234, or letters might be reversed within the word before shifting. Always check two or three letters before committing to a rule — one data point can mislead.
Memorise: E=5, J=10, O=15, T=20, Y=25. These five letters divide the alphabet into equal chunks. From any anchor, count forwards or backwards to find any letter's position instantly. Finding R? It is T(20) − 2 = 18. Finding L? O(15) − 3 = 12. Standard method (counting A=1, B=2... up to the letter): 8–12 seconds. EJOTY anchor method: 2–3 seconds for any letter.
Never test the shift on all letters. Take only the first two letters of the given word and the first two letters of the coded word, calculate both gaps, and confirm they match. If they do, the rule is confirmed and you apply without testing the rest. This cuts verification from 6 steps (one per letter) to 2 steps. For a 6-letter word, that is 4 fewer calculations — saving approximately 15–20 seconds per question.
When you get a 3-sentence coded language question, draw a quick 2-column table on your rough sheet. Left column: sentences. Right column: their codes. Visually scan for repeated words across rows — your eye spots them faster than sequential reading. Then match repeated codes. This structured layout prevents confusion between "which code belongs to which sentence" — the most common error in this type. Reduces error rate from roughly 30% (mental tracking) to near 0% (visual table), with no time added.
In operator-substitution questions, write the complete decoded arithmetic expression on your rough sheet before doing any calculation. Candidates who decode and calculate simultaneously frequently apply BODMAS in the wrong order under exam pressure. Writing it out first takes 4 extra seconds but eliminates the most common error in this subtype. Compared to mental calculation where wrong-order mistakes cost you the entire question, this is a net time saving of the 40–60 seconds wasted on re-attempting.
Some questions use mirror coding: A↔Z, B↔Y, C↔X... The rule is Position(letter) + Position(code) = 27. If you suspect mirror coding, check: does the first letter of the word plus the first letter of the code add to 27? If yes, confirmed in one step. Example: A(1) + Z(26) = 27. M(13) + N(14) = 27. Standard confirmation (listing all pairs): 30+ seconds. Formula check: 5 seconds.
When you see a Coding-Decoding question in the exam hall, run this decision tree in sequence:
Step 1 — Identify the type. Are codes symbols/words for whole words? → Type 2 (Coded Language, use elimination). Are codes operators? → Type 3 (BODMAS, write expression first). Are codes letters for letters? → Type 1 (Letter shift, check first two letters for gap). Are codes numbers for words? → Type 4 (Number-word, use elimination).
Step 2 — Find the rule. For letter shift: calculate gap from two letters only. For coded language: find words repeated across sentences, match their codes.
Step 3 — Verify on one more data point. Take a third letter or word to confirm the rule before applying.
Step 4 — Apply mechanically. Do not second-guess. Once the rule is confirmed, apply it to every element. Check your answer against the options before marking — a wrong alphabetic position or missed BODMAS step will match a distractor option.
Total time target per question: 45–60 seconds.
Why this question: Tests whether you can handle BODMAS under the pressure of decoded operators — the single most common error point in Type 3 questions.
Solving path: Step 1 — decode operators: find=+, for=÷, joy=×, new=−. Write: 8 + 16 ÷ 4 × 3 − 2. Step 2 — BODMAS: Division first: 16 ÷ 4 = 4. Then multiplication: 4 × 3 = 12. Then left to right: 8 + 12 = 20, then 20 − 2 = 18. Answer: 18.
Why this question: Classic 3-statement coded language question. Tests the elimination method with symbols — common in IBPS RRB.
Solving path: List: Statement 1 = [Send, money, home] → [@, #, ]. Statement 2 = [money, order, arrived] → [#, $, %]. Statement 3 = [home, loan, sanctioned] → [, &, +]. Find repeats: 'money' in S1 and S2 with '#'. So money=#. 'home' in S1 and S3 with ''. So home=. In S1, with money(#) and home(*) identified, only 'Send' and '@' remain. Answer: @.
Why this question: Tests +1 letter shift — the most frequent shift variant in IBPS RRB Clerk history.
Solving path: Check FLOWER→GMPXFS. F→G(+1), L→M(+1). Confirmed +1 shift. Apply to NATURE: N→O, A→B, T→U, U→V, R→S, E→F. Answer: OBUVSF.
Why this question: Tests −1 shift and your ability to apply it to a longer word without error.
Solving path: Check MOTHER→LNSGDQ. M→L(−1), O→N(−1). Confirmed −1 shift. Apply to SISTER: S→R, I→H, S→R, T→S, E→D, R→Q. Answer: RHRSDQ.
Why this question: Number-coded sentence type — appears when you need to match number codes to words through elimination. Slightly harder than symbol type because numbers look arbitrary.
Solving path: S1: [health, is, wealth] → [318, 49, 67]. S2: [wealth, brings, happiness] → [67, 28, 44]. S3: [happiness, is, temporary] → [44, 49, 56]. 'wealth' in S1 and S2 → code 67. 'happiness' in S2 and S3 → code 44. 'is' in S1 and S3 → code 49. In S2: wealth(67) and happiness(44) identified, only 'brings' and '28' remain. Answer: 28.
Stopping at one-letter confirmation for shift coding. You check F→G and assume +1, but the actual rule might be different for later letters (e.g., alternating shifts). Always verify with a second letter before applying.
Wrong BODMAS order in operator-substitution questions. The most costly mistake. Candidates add before dividing or multiply before dividing from left to right. Write the full expression, circle division and multiplication first, then handle addition and subtraction.
Confusing +1 and −1 shifts when the coded letters are unfamiliar. Under time pressure, G→H and H→G feel the same. Check direction explicitly: is the coded letter ahead or behind the original in the alphabet?
In coded language questions, assuming word order matches code order. "Send money home" coded as "@ # *" does not guarantee Send=@ unless you eliminate the other two words. Word order may or may not be preserved. Always use elimination, never assume position.
Forgetting wrap-around in shift coding. Z+1 = A, not AA or some invented letter. Similarly, A−1 = Z. Exam questions deliberately include letters near the ends of the alphabet to catch candidates who do not account for the circular alphabet.
Re-reading the question instead of the decoded expression. After replacing coded words with operators, some candidates glance back at the original question (with coded words) instead of their decoded expression. Work from your decoded expression only.