Coding–Decoding (New Pattern) — IBPS PO Reasoning

Forget the SSC-style "shift each letter by 3" puzzle — IBPS almost never asks it. The bank pattern is coded language / sentence coding: you're given a few sentences and their codes, and you must work out which code word stands for which real word by cross-referencing common words across sentences. It's a small logic puzzle, not a cipher. There are 3–5 of these per paper, and the intersection method cracks them cleanly every time. A handful of older forms (letter-number coding, symbol coding) still appear as one-offs, so this chapter covers both.

1. What IBPS actually asks

• Sentence / coded-language coding (the staple): "In a code, 'bank gives loan' is written '@ # $' and 'gives money fast' is '# % &'…" — deduce each word's code.
• Chinese/new-pattern coding: codes are assigned per word with no positional logic; pure intersection.
• Occasional older forms: letter-to-letter shift, letter-to-number, and symbol substitution — solvable by finding the rule (see §5).

The dominant skill is deduction across overlapping sentences, not memorising a shift.

2. The intersection method (this is the whole staple pattern)

Given several coded sentences:

1. Find a word common to two sentences. The code that appears in both their code-sets must stand for that common word.
2. Eliminate that pair from both sentences; repeat with the next common word.
3. A word appearing in only one sentence maps to the code appearing in only that sentence (once shared codes are removed).
4. Cross-check the final mapping against every sentence — it must be consistent throughout.

$\text{code(common word)}=\text{the code present in BOTH sentences’ code-sets}$

Do not try to guess by word order — new-pattern codes are deliberately position-scrambled. The intersection of code-sets is the only reliable signal.

3. Worked deduction

Codes:

• "bank gives loan" → pa ta ni
• "gives money fast" → ni ka la
• "loan for money" → ta ka za

Step 1 — common word 'gives' (sentences 1 & 2): shared code between {pa,ta,ni} and {ni,ka,la} is ni ⇒ ni = gives. Step 2 — common word 'loan' (sentences 1 & 3): shared between {pa,ta,ni} and {ta,ka,za} is ta ⇒ ta = loan. Step 3 — common word 'money' (sentences 2 & 3): shared between {ni,ka,la} and {ta,ka,za} is ka ⇒ ka = money. Step 4 — leftovers: sentence 1 leftover pa ⇒ bank; sentence 2 leftover la ⇒ fast; sentence 3 leftover za ⇒ for.

Now "bank money" = pa ka, and you can answer any question the set throws.

4. Handling ambiguity (two commons, no unique intersection)

Sometimes two sentences share two words, so the intersection gives a pair of codes for a pair of words without separating them. Don't force it — bring in a third sentence containing just one of those two words; its overlap breaks the tie. If the set genuinely can't resolve a specific pair, the questions are designed so you never need that exact word — answer what is determined.

5. The older forms (one-off insurance)

• Letter shift: find the constant gap. CAT → ECV is +2 each letter; decode by shifting back.
• Letter–number: A=1 … Z=26 (or reverse A=26). BAD → 2,1,4. Watch for "sum of positions" or "product" variants.
• Symbol/conditional coding: a table of digit→symbol with rules ("if the first and last are both odd, swap their codes"). Apply the conditions in order; the trap is a condition you skipped.
• Mirror/opposite: A↔Z, B↔Y (opposite letter = 27 − position).

Quick tool: the EJOTY landmarks — E=5, J=10, O=15, T=20, Y=25 — let you place any letter's number in seconds without counting from A.

6. Solved examples

Q1 (intersection). "he is tall" = sa re na; "she is short" = re pi tu. What is the code for "is"? Solution. Common word 'is'; shared code between {sa,re,na} and {re,pi,tu} = re. re = is.

Q2 (letter shift). If DELHI is coded GHOKL, then MUMBAI is? Solution. Shift +3 (D→G, E→H…). M→P, U→X, M→P, B→E, A→D, I→L = PXPEDL.

Q3 (letter–number, EJOTY). If LAMP → 12,1,13,16, then WORD → ? Solution. W=23 (Y=25, so W=25−2), O=15, R=18, D=4 ⇒ 23,15,18,4.

Q4 (opposite letter). In a code A→Z, B→Y…, CAB becomes? Solution. Opposite = 27 − position: C(3)→X(24), A(1)→Z(26), B(2)→Y(25) = XZY.

7. The protocol

For a sentence-coding set (the common case):

1. List each sentence's code-set.
2. Match common words to common codes by intersection; eliminate and repeat.
3. Assign leftovers and cross-check the full mapping against all sentences.
4. Answer the questions from the finished dictionary.

For an older-form one-off: find the single rule (shift / position / opposite / condition), verify it on the given example, then apply to the target. Use EJOTY to place letter-numbers fast.