Rational or Irrational? Hunt for the Repeating Block

You get a long decimal on the board — say 0.142857142857142857\dots — and the question is: rational or irrational? You could try to find a fraction for it. You could plug it into a calculator. Or you could use the one-line mental test that a student trained on this stuff uses on autopilot: scan for a repeating block of digits. If a block of digits eventually locks in and keeps repeating, the number is rational. If it doesn't, suspect irrational.

This satellite is about how to run that scan quickly and when it works.

The theorem behind the trick

Every rational number \dfrac{p}{q} has a decimal expansion that either terminates or eventually repeats. And the converse is also true — every terminating or eventually-repeating decimal is rational. So the problem of "is this number rational?" becomes the problem of "does this decimal eventually fall into a repeating pattern?"

Why the theorem holds: when you long-divide p by q, every step produces a remainder that is an integer between 0 and q - 1. That is only q possible remainders. After at most q steps you must either hit a remainder of zero (terminating) or see a remainder you've seen before (which forces the digits from that point on to repeat the pattern you already computed). There is no third option.

A terminating decimal like 0.25 is really an eventually-repeating decimal where the repeating block is \overline{0}: write 0.25 as 0.25000000\dots and the block 0 repeats forever. So you can reduce everything to "find the repeating block."

The scan — step by step

When you stare at a decimal, run this scan:

Does it terminate? If the digits stop after finitely many places, it is rational. Done.
Is a single digit repeating forever? Scan the tail. Do you see \dots 3333\dots or \dots 6666\dots? If yes, rational with period 1.
Is a two-digit block repeating? Look for patterns like \dots 272727\dots or \dots 818181\dots. The same two digits alternating.
Is a longer block repeating? For \tfrac{1}{7} the block is 142857 — six digits long. Scan the last 20–30 digits of the decimal and look for a chunk that repeats.
If no block appears up to length, say, 10–15, suspect irrational. Famous irrationals (\pi, e, \sqrt{2}, \sqrt{3}) have decimal tails with no repeating block at any length.

Top: $\tfrac{1}{4}$ terminates, so the repeating block is $0$. Middle: $\tfrac{1}{7}$ has period-$6$ block $142857$. Bottom: $\sqrt{2}$ shows no block even in the first eleven digits — and never will.

Worked mental passes

Mental pass A: 0.818181818181\dots

Length-2 block 81 is screaming at you from the first digit. Rational. If you want the fraction, use the 100x - x trick: x = 0.818181\dots, 100x = 81.818181\dots, so 99x = 81, giving x = \dfrac{81}{99} = \dfrac{9}{11}. But you didn't need that to answer the rational question.

Mental pass B: 0.1010010001000010000010000001\dots

Dangerous. It looks patterned — there is clearly a rule (1 followed by 0, 00, 000, 0000, \dots). But the pattern is not a repeating block. The number of zeros between 1s keeps growing. No finite block of digits ever locks in. So the decimal is irrational. (It is a well-known irrational, called Liouville-style by construction.)

Why patterned does not mean rational: "repeating" has a strict meaning — the same finite block, over and over, with no variation. A pattern that grows or shifts does not count. Your scan must look for exact repetition, not any kind of structure.

Mental pass C: 3.14159265358979323846\dots

You recognise \pi. Even if you didn't, no block of length up to 10 repeats. Suspect irrational. (In fact \pi is known to be irrational — proved by Lambert in 1761.)

How far should you scan before giving up?

Technically, if a rational has denominator q, its period is at most q - 1 digits. So to be certain a decimal is rational by scan alone, you would need to scan up to tens of thousands of digits for large denominators — which is impractical by hand.

In practice, for exam problems:

If you recognise the decimal as one of \pi, e, \sqrt{2}, \sqrt{3}, \sqrt{5}, \dots, the answer is given by what you know, not the scan.
If the decimal is presented with a bar (0.\overline{27}, 0.1\overline{6}), the block is explicitly marked — it is rational.
If the first 8–10 digits clearly cycle, declare rational.
If the problem supplies a construction rule ("each digit is 1 if the place number is prime, else 0"), scan for repetition in the rule itself, not in the digits. Most constructed decimals that generate non-periodic patterns are irrational.

Applying the scan under exam pressure

A Class 9 MCQ asks: "Which of the following is irrational?"

(a) 0.123123123\dots (b) 0.12121212\dots (c) 0.1234567891011121314\dots (Champernowne-style — you just list the counting numbers) (d) 0.\overline{142857}

Scan each:

(a): block 123 repeats. Rational.
(b): block 12 repeats. Rational.
(c): the "rule" is deterministic, but the digit sequence itself shows no repeating block — the counting numbers keep growing and leaking new digit patterns into the tail. Irrational.
(d): bar marks 142857 as the block. Rational (in fact \tfrac{1}{7}).

Answer: (c). The scan caught it in seconds. You did not need to do any arithmetic.

The misconceptions to avoid

Three traps that students fall into with this scan:

"Non-terminating means irrational." No. 0.333\dots is non-terminating and rational. The condition is non-terminating AND non-repeating.
"If the decimal has a pattern, it's rational." No. A decimal like 0.101001000100001\dots has a clear construction pattern but no repeating block. It is irrational.
"I need to compute the fraction to prove rational." No. The scan alone (spot a repeating block) tells you rationality. Converting to a fraction is a separate problem, solved by the 10^n x - x trick.

The scan is a fast triage tool. It decides the rational-vs-irrational question in a few seconds, so you can spend your time on the parts of the problem that actually need computation.