ExamShortcut
high importance~3 Q in Tier 19 formulas⚡ 11 shortcuts6 subtopics

Mixed clusters and same-relation selection

🔒 Log in to track

Two CGL variants sit outside the classic three:

  1. Mixed cluster analogy: A : B :: C : ? where A and B contain both letters and numbers ('B3D : E6J'). Each component follows its own mini-rule (letters shift, number doubles, etc.). Solve component-wise and assemble the answer.
  2. Same-relation selection: four complete pairs are given; pick the pair whose relation matches a stated model pair (or pick the pair that does NOT belong, when asked). Test all four pairs with the same relation sentence from the model.

For mixed clusters, never solve the cluster as a whole — split by character type and solve each track separately.

Detailed notes

What is a mixed analogy?

Mixed analogies join two worlds: letters and numbers, or words and numbers. The good news is that each part follows its own simple rule. You solve the parts one at a time and then put the answer together.

Type 1: letter-number clusters

Example: C5 : F10 :: H8 : ?

  • Letter track: C(3) → F(6) is +3. So H(8) + 3 = K.
  • Number track: 5 → 10 is × 2. So 8 × 2 = 16.
  • Answer: K16.

The two tracks do not affect each other. Solve the letters with positions (EJOTY helps) and the numbers with arithmetic. If a cluster has two letters (B3D), each letter slot has its own shift.

Sometimes the number is built from the letter: D16, F36, H64 — each number is the square of its letter's position (D = 4, 4² = 16). Then the next letter decides the number.

Type 2: word → number

Example: CAT : 24 :: BAT : ? The number comes from the letter positions of the word. The rule bank:

  1. Sum of positions: CAT = 3 + 1 + 20 = 24.
  2. Sum × k or sum ± r: DOG = 26, coded as 52 → × 2.
  3. Reverse positions (A = 26, B = 25 ... Z = 1): each letter counts as 27 − p.
  4. Number of letters × something (rare).

Work out the sum for the model word first; then compare it with the given number to find the extra step.

Type 3: letter → number

Example: C : 9 :: E : ? C is the 3rd letter and 9 = 3². So E (5th) gives 5² = 25. Other common rules: 2 × position, reverse position (B → 25), position + constant.

Type 4: same-relation word pairs

"Select the pair that is related in the same way as Tailor : Needle." Say the link — worker : tool — and test all four pairs. Traps are pairs from the same world with a different link: Doctor : Hospital (workplace), Author : Book (product).

Solving method in short

  1. Split the term into its parts (letters, numbers, word, number).
  2. Find one small rule per part from the model pair.
  3. Apply each rule to the matching part of the third term.
  4. Join the parts and look for the exact combination in the options (traps change just one part).

Common mistakes

  • Solving the letter part correctly but copying the wrong number (or the reverse). Options usually differ in only one part.
  • Using A = 0 instead of A = 1 when counting positions.
  • For word → number, forgetting the extra × k or + r that the model pair revealed.
  • In pair selection, choosing a pair from the same topic instead of the same link.

Quick revision

  • Split, solve each track, then join.
  • Letter tracks: positions and shifts. Number tracks: +, ×, squares.
  • Word → number: first try the plain sum of positions, then × k or ± r.
  • Letter → number: position, 2 × position, position², reverse position (27 − p).
  • Pair selection: one link sentence tested on all four pairs.

Types of questions asked

Every way this subtopic shows up in exams — how to recognise it, the formula or logic to use, and a solved example.

Type 1: Letter-number cluster analogycommon3 practice Q
How to spot it:

Terms like 'C5 : F10' or 'P7R : S14U' mix letters and numbers.

  1. Split each term into a letter track and a number track.
  2. Find the letter shift and the number rule separately.
  3. Apply both to the third term and join the parts.

Why it works: the tracks are independent, so each one is a tiny, easy question.

Example: C5 : F10 :: H8 : ?

Letters: C → F is +3, so H → K. Numbers: 5 → 10 is × 2, so 8 → 16. Answer: K16.

Type 2: Word to number (sum of letter positions)common3 practice Q
How to spot it:

A word is paired with a number, e.g. 'CAT : 24 :: BAT : ?'.

code=k∑pi+r\text{code} = k\sum p_i + r
  1. Add the positions of the model word's letters.
  2. Compare the sum with the given number: equal, × k, or ± r.
  3. Find the sum for the new word and apply the same extra step.

Why it works: setters build the number from letter positions in one or two simple steps.

Example: RAM : 32 :: SHAM : ?

R + A + M = 18 + 1 + 13 = 32 (plain sum). SHAM = 19 + 8 + 1 + 13 = 41.

Type 3: Select the word pair with the same relationcommon3 practice Q
How to spot it:

A model pair followed by four complete word pairs: 'Select the pair related in the same way as Tailor : Needle'.

  1. Write the link of the model pair as a sentence.
  2. Put each option pair into that sentence.
  3. Only one pair should make a true sentence in the same direction.

Why it works: traps share the topic but not the link (workplace instead of tool).

Example: Tailor : Needle — (Painter : Brush / Doctor : Hospital / Author : Book / Farmer : Crop)

Link: worker : tool used. Painter uses a brush ✓. Hospital = workplace, book and crop = products. Answer: Painter : Brush.

Type 4: Letter to number (position-based)occasional3 practice Q
How to spot it:

A single letter is paired with a number, e.g. 'C : 9 :: E : ?' or 'B : 25 :: D : ?'.

n=p2,  2p,  27−pn = p^2,\; 2p,\; 27-p
  1. Write the letter's position (and its reverse position 27 − p).
  2. See how the number is built: p², 2p, 27 − p, p + r.
  3. Apply the same build to the new letter.

Why it works: only a few simple formulas are used, and each is quick to test.

Example: C : 9 :: E : ?

C is 3rd and 9 = 3². E is 5th, so 5² = 25.

Shortcut tricks

⚡ Solve tracks, then assemble

In a mixed cluster make one column for letters (positions/shifts) and one for numbers (their own series). Fill the answer character by character — each column is independent, so an error stays inside one column.

Example: Q. 'A5C' : 'D10F' :: 'G15I' : ?

Sol. Letters: A→D (+3), C→F (+3); numbers: 5→10 (×2). Apply: G+3=J, I+3=L, 15×2=30 → J30L.

Component-wise solving turns 1 hard cluster into 2 easy columns.

⚡ One sentence for all four pairs

In same-relation selection, form the relation sentence once from the model pair and tick/cross every option against the same sentence, in the same order. The first option that passes both halves is the answer.

Example: Q. 'Sword : Sheath' :: pick the same-relation pair — (a) Rifle : Trigger (b) Knife : Blade (c) Rose : Petal (d) Blade : Cover is not; use (d) Bullet : Barrel

Sol. Sentence: 'second is the container/cover of the first when stored'. Trigger, Blade, Petal are parts, not containers. Bullet : Barrel — barrel houses the bullet.

Container–thing, worker–tool, cause–effect: name the abstract frame, then match frames.

Where students lose marks

  • Assuming the number track follows the letter track's rule — the two tracks are independent.

  • In same-relation selection, matching only the first half of the relation (both 'container' pairs) without checking the second half (what is contained).

  • Missing an embedded digit rule like 'number = position of the first letter' (B3D → 3 is C's position... actually A=1; check the model, don't assume).

Practice sets — 13 questions

Sets of 10, mixed across the question types above. Each answer comes with a step-by-step explanation.

Topic test · 13 questions

Suggested time 10 min · wrong answers go to your mistake notebook automatically.