Non-Verbal Reasoning: Figures, Mirrors and Series

By the end of this lesson you will be able to look at a set of figures, name the kind of question it is, find the single rule behind it, and pick the exact figure that the rule demands.

Already confident? Skip to practice: Non-Verbal Reasoning quiz

Before this: no special facts are needed, but the habit that figure questions reward is the same one the Number and Letter Series lesson builds — find the rule, then apply it once more. Figure series are that lesson drawn as shapes instead of numbers.

These questions are visual

Non-verbal reasoning questions are pictures. Each one shows you figures — shapes, patterns, dots, arrows, part-filled boxes — and asks you to choose the figure that comes next, matches, or breaks the pattern. The quiz shows them to you as images, and you solve them by looking, not by reading a description.

That is why this lesson has no figures to read. Figures on a page here would only be one more example to memorise. Instead it teaches you how to see the pattern: what to look at first, which one thing to track, and the traps that make two answers look alike. When you open the quiz, you bring this way of looking to the real pictures.

So everything below is described in words. Read each method, picture the figures in your head, and then recognise the same move when you meet it as an image. The worked examples are also described in words — read them slowly, draw them on scrap paper if it helps, and the shapes they describe will be easy to recognise once they are in front of you as images.

One habit carries across every type: look for the single thing that changes, and ignore everything that stays the same. A figure question almost never has two unrelated rules at once; find the one rule, and the exact answer follows.

The main types you will meet

SSC and CHSL non-verbal questions come in a handful of shapes. Learn to name the one in front of you first — the method follows from the name.

Figure series and rotation

A series shows three or four figures in a row that change by one steady rule, then asks for the figure that comes next.

The whole skill is finding what changes each step and leaving alone what does not. The common changes are:

• Rotation — the whole figure, or one arm or marker on it, turns by a fixed angle each step.
• An element added — one more dot, line or side appears at each step.
• A part moving — a shaded cell, an arrow or a small shape moves around the corners or edges, one position each step. A shaded square creeping clockwise around the four corners of a grid is the classic case.

For rotation, two things decide the answer: the direction and the size of the turn. Pick one clear marker on the figure — an arrow head, a single spoke, the open side of a shape — and watch only that marker. If it moves from pointing up to pointing right, that is a quarter turn (90°) clockwise; up to left is 90° anticlockwise; up to down is a half turn (180°), where direction no longer matters. The common turns are 45° (one-eighth), 90° (a quarter) and 180° (a half); measure against the clock face — twelve to three is a quarter, twelve to six is a half. Measure the turn between the first two figures, confirm it is the same between the next two, then turn the last figure by that same amount in that same direction to get the answer.

When something is also being added or moved, track each change on its own line: "the triangle turns 90° clockwise and one dot is added." Apply every change once more, together, to the last figure.

A worked example, in words. Picture a square with one short arrow inside it. In the first figure the arrow points up; in the second it points right; in the third it points down. What does the fourth figure show?

1. Track one marker — the arrow head. Up to right is a quarter turn clockwise. Right to down is another quarter turn clockwise. The rule is "turn 90° clockwise each step".
2. Your turn: turn the arrow one more quarter turn clockwise from "down".

Check: down turns to left. The fourth figure is a square with the arrow pointing left.

A second worked example — a part moving. Picture a 2×2 grid of four cells. In the first figure the top-left cell is shaded; in the second, the top-right; in the third, the bottom-right. What does the fourth figure show?

1. Track the shaded cell. It moves top-left → top-right → bottom-right — one step clockwise around the corners each time.
2. Your turn: move it one more step clockwise.

Check: the next corner clockwise is bottom-left, so the fourth figure has the bottom-left cell shaded. Here nothing rotates or is added; the only change is the shaded cell walking round the grid, and you apply exactly that.

Mirror image vs water image

This is the type candidates most often get backwards, so fix the two definitions firmly and do not swap them.

• A mirror image is a left–right flip. Imagine a mirror standing upright beside the figure, along its right or left edge. Left and right swap; top and bottom stay put. The letter b becomes d; a shape leaning right now leans left.
• A water image is a top–bottom flip. Imagine still water below the figure, as on a lake. Top and bottom swap; left and right stay put. The letter b becomes p; a shape standing up now hangs down.

A quick test for either one: find a feature that is clearly on one side. For a mirror, ask "does it move from left to right (or right to left) while staying at the same height?" For water, ask "does it move from top to bottom while staying on the same side?" If only the horizontal has swapped, it is a mirror; if only the vertical has swapped, it is water.

Your turn, in words. The digit 6 stands upright. Picture its water image — the reflection you would see in a lake below it. Which way does it turn?

Check: the water image flips top and bottom, so the loop of the 6 moves to the top and the figure reads like a 9 leaning the same way — upright left-to-right, but turned over. (Its mirror image, by contrast, would keep the loop at the bottom and only face the other way.)

A worked example, in words. Take the capital letter R standing upright. Its mirror image, flipped left–right, is a backwards R — the vertical stroke moves to the right side and the loop opens to the left, but it still stands the right way up. Its water image, flipped top–bottom, is an upside-down R — the loop and leg now hang from the top, but left and right are unchanged. Notice how different the two are: the mirror keeps it upright, the water turns it over.

The classic trap is reading the word — "mirror" or "water" — too fast and flipping the wrong way. Mirror is the one you stand next to, so it flips sideways; water is the one below you, so it flips up–down. Say that to yourself before you choose.

Odd one out (classification)

A classification question shows four (sometimes five) figures and asks which one does not belong. Three of them share a property; the fourth breaks it.

Work by hunting for the shared property, not by staring at the odd one. Check these, roughly in order:

• Shape — three are closed curves and one has a corner; three are made of straight lines and one has a curve.
• Fill — three are shaded and one is blank, or three have the same part shaded.
• Number of sides or elements — three have an even number of lines, three have the same count of dots.
• Rotation or symmetry — three can be turned onto each other, one is a mirror of the rest.

A worked example, in words. Four figures: a triangle, a square, a pentagon and a circle. Three of them are made only of straight lines; the circle is not. The odd one out is the circle, because the shared property is "a shape made of straight sides" and the circle alone breaks it.

Your turn, in words. Four figures, each a circle with an arrow inside: in three of them the arrow points outward, away from the centre; in one it points inward, toward the centre. Which breaks the pattern?

Check: the shared property is "arrow pointing outward". The odd one out is the figure whose arrow points inward. Note that the shape (a circle with an arrow) is the same in all four — the rule lives in the direction, not the outline, which is why you must check fill, count and direction, not just shape.

The moment you find a property that exactly three figures share, the fourth is your answer. If two different properties each single out a different figure, prefer the simplest, most obvious one — the exam intends a single clean rule.

Analogy (A : B :: C : ?)

An analogy gives you a pair, A and B, that are related by some change, then a third figure C, and asks for the figure that stands to C as B stands to A.

1. Work out the transformation from A to B — what was done to A to make B? It might be a rotation, a mirror or water flip, adding or removing an element, or shrinking part of the figure.
2. State that change in words: "B is A turned 90° clockwise with one extra dot."
3. Apply the same change to C, and choose the option that matches exactly.

A worked example, in words. A is an upward arrow; B is the same arrow with a small circle added at its tail. C is an upward triangle. What is the fourth figure?

1. The change from A to B is "add a small circle at the tail" — nothing is rotated or flipped.
2. Your turn: do the same to C.

Check: the answer is an upward triangle with a small circle added at its tail. Apply only the change you saw — do not also rotate C, because A was never rotated to make B.

Sometimes the change from A to B is two things at once — a rotation and a removed element, or a flip and a shrink. Name both, then apply both to C. The safest check is to go through the options and reject any that gets even one part wrong; the one that survives is the answer.

The skill is the same "find the rule, apply it" habit as a series — the rule just lives in the first pair instead of across a row.

Counting and embedded figures

Two related types ask you to count or to find a figure hidden inside another.

For a counting question ("how many triangles are in this figure?"), do not count at random — you will miss some and double-count others. Count systematically: first every smallest single triangle, then the ones made of two smaller pieces joined, then three, and so on, keeping a running total. A system is the only way to be sure you have them all.

A worked counting example, in words. Picture a large triangle with a single line drawn from its top corner straight down to the middle of the base, splitting it into two. How many triangles in all?

1. Smallest pieces first: the line makes two small triangles, left and right.
2. Combinations next: the left and right pieces together form the one big original triangle.
3. Running total: 2 + 1 = 3 triangles. Counting at random, most people stop at 2 and miss the big one — the system catches it.

For an embedded figure question, you are given a simple shape and asked which of the four complex figures contains it, hidden, in the same size and orientation. Fix the exact outline of the given shape in your mind, then look in each option for that outline sitting inside the lines — not merely a similar shape, but that exact one, unturned and unresized.

For example, if the given shape is a right-angled triangle, you are not looking for any triangle in the options — you are looking for that triangle, the same size and the same way up, traced by lines that are already in the figure. A triangle of the right shape but tilted, or larger, does not count.

A fast way to narrow the options: find one unusual feature of the given shape — a sharp spike, a particular notch, a right angle where two lines meet — and scan the four figures for that feature. Any option that cannot even contain that one feature is out, and you need only look closely at the rest.

How to attack any figure question

1. Name what kind it is — series, mirror or water image, odd one out, analogy, counting, or embedded. The method follows from the name.
2. Look for the one thing that changes (and note what stays the same). Pick a single marker to track rather than watching the whole figure at once.
3. State the rule in words — "turns 90° clockwise and adds one dot", "flipped left–right", "three are shaded, one is blank".
4. Apply the rule once more, or test each option against it. Reject any option that gets even one detail wrong.
5. Watch mirror-vs-water and the rotation direction. Mirror flips left–right; water flips top–bottom; confirm clockwise or anticlockwise before you turn.

At a glance: the types in one place

Name the type first, then find the single rule and apply it exactly.

• Figure series and rotation: find what changes each step — rotation by a fixed angle, an element added, or a part moving round the corners — and apply it once more. Track one marker; fix the direction and size of any turn.
• Mirror image: a left–right flip, as in a mirror standing beside the figure. Left and right swap; top and bottom stay.
• Water image: a top–bottom flip, as in a reflection in water below. Top and bottom swap; left and right stay.
• Odd one out: find the property three figures share — shape, fill, number of sides, rotation — and the fourth that breaks it.
• Analogy (A : B :: C : ?): work out the change from A to B, then do the same to C.
• Counting figures: count systematically — all the smallest first, then combinations — keeping a running total.
• Embedded figures: fix the given shape's outline, then find it hidden, same size and orientation, inside one option.

Common mistakes

• Confusing mirror and water images. Mirror is left–right (the mirror stands beside you); water is top–bottom (the water is below you). Flipping the wrong way is the single commonest error in this topic.
• Guessing the rotation direction. Decide clockwise or anticlockwise from one clear marker, and check the angle is the same between every pair, before you turn the last figure.
• Assuming a change that is not there. Track only what actually changes from figure to figure; do not invent an extra rotation or an added dot the pattern never shows.
• Picking the figure that is "close". In figures the answer is exact. An option with the right shape but the wrong shading, one missing line, or a slight turn is wrong — check every detail against your rule.
• Counting at random. Without a system you will miss figures or count one twice. Count the smallest pieces first, then the combinations.

Next

The other Reasoning topics — Number and Letter Series, Coding–Decoding, Syllogism, Blood Relations and Direction Sense — build the same "find the rule, apply it" habit on words and numbers; this topic is that habit applied to pictures. Because these questions are visual, the only way to get fast is to see many of them: with every figure you work, naming the type and spotting the one change gets quicker, until a glance is enough.

The quiz is where this lesson pays off — the figures there are the real images these methods were written for. Take it with scrap paper beside you, name each type before you solve, and track the single thing that changes.

Ready to test this lesson? Take the Non-Verbal Reasoning quiz

How this was made: Drafted with AI assistance; every answer re-solved by a separate check that had not seen the key; spot-checked by the editor on 29 October 2026.