Dishantra Insights · Class 8 Mathematics

Number Play Class 8: Always, Sometimes or Never? 8 Divisibility & Remainder Traps

← Education, Skills & Policy

In Number Play, the hardest questions are often not long calculations. They are short claims that look true. The useful skill is deciding whether a statement is always true, sometimes true or never true—and knowing whether you need a proof or just one counterexample.

This guide does not reproduce NCERT exercise solutions. It uses independently created examples to show eight recurring traps in divisibility, remainders and number structure, followed by a changed question to test whether the reasoning transfers.

First choose the right test: proof or counterexample?

For an always claim, examples are not enough—you need a reason that works for every allowed number. To show a claim is not always true, one valid counterexample is enough. A sometimes claim needs at least one case where it works and one where it fails.

ALWAYS?
Write a general form such as 6a or 5k + r, then prove the property.
SOMETIMES?
Find one example that works and one counterexample that fails.
NEVER?
Show the structures cannot overlap—for example odd versus even.
REMAINDER?
Translate “remainder r on division by d” into dk + r.
Dishantra original instructional graphic · Name the claim type before calculating.
1

“If 6 divides two numbers, it divides their sum.”

Verdict: Always true. If the numbers are 6a and 6b, their sum is 6a + 6b = 6(a+b), which is still a multiple of 6.

Why examples alone are not the proof

24 + 42 = 66 works, but one working example cannot establish “always”. The algebraic form does.

Changed check: If 7 divides x and y, must 7 divide x−y? Yes: x=7a, y=7b, so x−y=7(a−b).

2

“If 6 divides a sum, it must divide both addends.”

Verdict: Sometimes true, not always. 12 + 18 = 30 and 6 divides both addends. But 1 + 5 = 6 even though 6 divides neither 1 nor 5.

Trap: This reverses a true rule. From “if A, then B” you cannot automatically conclude “if B, then A”.

Repair

When you reverse a divisibility statement, test the converse with a small counterexample before trusting it.

Changed check: 8 divides 3 + 5. Does 8 divide either addend? No.

3

“Divisible by 4 and by 6 means divisible by 24.”

Verdict: Sometimes true. A number divisible by both 4 and 6 is guaranteed to be a multiple of LCM(4,6)=12, not necessarily 24.

12 passes both the 4-test and the 6-test but fails the 24-test. Meanwhile 48 passes all three.

Repair

For “divisible by both” questions, combine the prime-factor requirements. 4 = 2² and 6 = 2×3, so the least requirement is 2²×3 = 12.

Changed check: Divisible by 8 and 3 guarantees divisibility by what least number? 24.

4

“If the digit sum is divisible by 9, the number is divisible by 9.”

Verdict: Always true for whole numbers written in base 10. For example, 73,458 has digit sum 7+3+4+5+8 = 27, so it is divisible by 9.

Why it works: each place value 10, 100, 1000, … leaves remainder 1 when divided by 9, so the number and its digit sum leave the same remainder.

Wrong shortcut to avoid

Do not turn this into “any convenient digit operation works”. The shortcut depends on the divisor.

Changed check: Is 58,347 divisible by 9? Yes. Digit sum = 27.

5

“The digit-sum test for 9 also works for 11.”

Verdict: Sometimes true by accident, not as a rule. 29 has digit sum 11, but 29 is not divisible by 11.

The 11-test uses the alternating sum of digits, not the ordinary digit sum. For 2,904: (2+0) − (9+4) = −11, so 2,904 is divisible by 11.

Repair

Match the shortcut to the divisor. A correct shortcut copied to the wrong divisor becomes a new error.

Changed check: Is 4,862 divisible by 11? Yes: (4+6) − (8+2) = 0.

6

“Two numbers with the same remainder have a difference divisible by the divisor.”

Verdict: Always true. If two numbers leave remainder r on division by d, write them as da+r and db+r. Their difference is d(a−b).

For example, 38 and 68 both leave remainder 3 on division by 5. Their difference is 30, which is divisible by 5.

Remainder fingerprint

The shared remainder disappears on subtraction. That is often the fastest way to recognise the structure.

Changed check: 47 and 75 both leave remainder 5 when divided by 7. Is their difference divisible by 7? Yes: 28.

7

“Numbers leaving remainder 3 on division by 5 must look only like 5k + 3.”

Verdict: The structure is right, but the expression is not unique. 5k+3 is the standard form. Yet 5k−2 can describe the same remainder class because 5k−2 = 5(k−1)+3.

For k = 1,2,3,… the second form gives 3, 8, 13, …—the same pattern.

Repair

Judge an expression by the remainder it produces, not by whether it visually matches the first form you learned.

Changed check: Which has remainder 4 on division by 6: 6k+4 or 6k−2? Both.

8

“Odd + even can be a multiple of 6.”

Verdict: Never true. Odd + even is always odd. Every multiple of 6 is even. A number cannot be both odd and an even multiple of 6.

Repair

Before testing many values, compare the structural categories. Parity can settle the claim immediately.

Changed check: Can odd + odd be a multiple of 6? Sometimes: 1+5=6, but 1+3=4.

Remainder translation
DIVISOR
d
the group size
WHOLE GROUPS
dk
a multiple of d
LEFTOVER
+ r
0 ≤ r < d
FULL FORM
dk + r
same remainder class
Dishantra original instructional graphic · Translate words into structure before applying a shortcut.

Use the wrong claim as a clue to the missing idea

If a learner says “divisible by 4 and 6 means divisible by 24”

The likely gap is combining conditions by multiplication instead of by least common multiple / prime-factor requirements.

If a learner says “digit sum works for 11 too”

The learner may know one divisibility shortcut but not why it works. Route back to place-value remainders, not another memorised list.

If a learner cannot decide Always / Sometimes / Never

The gap may be proof strategy: they need to know when to use a general form, a counterexample, or a structural contradiction.

This is why a Number Play error can be more useful than a raw score. The wrong answer often identifies the decision rule that needs repair.

A compact four-step Number Play check

1. Translate the claim

Write multiples as dk and remainder statements as dk+r.

2. Ask what would prove it

General algebra for “always”; one counterexample can destroy “always”; two contrasting cases can establish “sometimes”.

3. Check the condition, not just the shortcut

Digit sum belongs to 3 and 9; alternating digit sums belong to 11; combined divisibility may require an LCM.

4. Change the numbers

If the rule survives a changed example, the learner is more likely to have the idea rather than the memorised answer.

Three transfer checks

Try these without copying the examples above.

Check 1: A number is divisible by both 6 and 10. What least number must divide it?

LCM(6,10) = 30.

Check 2: Two numbers both leave remainder 4 when divided by 9. What can you say about their difference?

Their difference is a multiple of 9, because (9a+4)−(9b+4)=9(a−b).

Check 3: “Even + even is a multiple of 4.” Always, sometimes or never?

Sometimes. 2+2=4 works, but 2+4=6 does not.

Sources & curriculum reference

Primary curriculum reference: National Council of Educational Research and Training (NCERT), Grade 8 Ganita Prakash, Part I, Chapter 5 Number Play (current 2026–27 edition; curriculum structure independently checked against current online chapter indexes).

All examples, counterexamples, transfer checks and diagrams on this page are independently created by Dishantra. This page does not reproduce NCERT exercise text and is not affiliated with or endorsed by NCERT, CBSE or any school.

Visual sources: both instructional graphics are original Dishantra diagrams; no external image assets are used.

Editorial note

Curriculum coverage and the current search landscape were reviewed on 28 October 2026. Current results contain large solution banks, worksheets and direct Always/Sometimes/Never answer pages, so this guide uses a diagnostic claim → proof/counterexample → repair → transfer structure rather than reproducing a solution set.

Prepared with AI-assisted research and writing, then independently rechecked for mathematical correctness, wording assumptions and changed-context transfer under Dishantra’s educational semantic-QA standard. Dishantra editorial standards