A classic problem: how many positive divisors does 720 have? A student who has not seen the trick starts listing: 1, 2, 3, 4, 5, 6, 8, 9, 10, 12, 15, 16, \ldots — and gets tangled, misses some, and double-counts others. A student who has internalised the trick answers in one line. The trick is to recognise that the question is not really about divisors — it is about choices of exponents.
Once you see that, "number of divisors" problems, "number of common divisors" problems, "smallest number with exactly N divisors" problems, and many Olympiad questions collapse into short exponent-counting arguments.
The core identity
Let n have the prime factorisation
Then the number of positive divisors of n, written \tau(n) or d(n), is
Why: a positive divisor of n has the form p_1^{b_1} p_2^{b_2} \cdots p_r^{b_r} where 0 \leq b_i \leq a_i. For each prime p_i you have a_i + 1 choices for b_i (the values 0, 1, 2, \ldots, a_i). The choices are independent, so by the multiplication principle, the total number of divisors is the product \prod (a_i + 1).
That one line of reasoning — "a divisor is a choice of exponent at each prime" — is what reframes the whole question.
The standard example
720 = 2^4 \cdot 3^2 \cdot 5. So
720 has exactly 30 positive divisors. You did not list a single one; you only looked at the exponents.
The recognition rule
You see any of these phrasings in a problem:
- "How many positive divisors does n have?"
- "How many factors of n are greater than k?"
- "Find n such that n has exactly 12 positive divisors."
- "Which n has the most divisors?"
- "How many common divisors do a and b share?"
- "For how many n \leq 100 is \tau(n) = 6?"
Your response should be automatic: factor everything into primes and turn the question into an exponent-counting question. That is the entire thinking-process skill. You will almost never need anything else.
Four variants you should recognise
Variant 1: "Divisors greater than k"
How many positive divisors of 120 are greater than 12?
120 = 2^3 \cdot 3 \cdot 5, so \tau(120) = 4 \cdot 2 \cdot 2 = 16. The divisors of 120 come in pairs (d, 120/d) whose product is 120. If d < \sqrt{120} \approx 10.95, then 120/d > 10.95. So the number of divisors > \sqrt{120} equals the number < \sqrt{120}.
Divisors \leq 10: 1, 2, 3, 4, 5, 6, 8, 10 — that's 8. So divisors \geq 12 = 16 - 8 = 8.
The trick of pairing divisors (d, n/d) is a direct consequence of the exponent-picking view: picking (b_1, b_2, b_3) = (\text{choices}) gives d, and picking (a_i - b_i) gives n/d.
Variant 2: "Numbers with exactly N divisors"
Find the smallest positive integer with exactly 12 divisors.
You want \tau(n) = (a_1 + 1)(a_2 + 1) \cdots = 12. Factor 12 as a product of integers \geq 2:
- 12 = 12. Then a_1 + 1 = 12, so a_1 = 11; smallest such n is 2^{11} = 2048.
- 12 = 6 \cdot 2. Then exponents 5, 1; smallest n = 2^5 \cdot 3 = 96.
- 12 = 4 \cdot 3. Exponents 3, 2; smallest n = 2^3 \cdot 3^2 = 72.
- 12 = 3 \cdot 2 \cdot 2. Exponents 2, 1, 1; smallest n = 2^2 \cdot 3 \cdot 5 = 60.
- 12 = 2 \cdot 2 \cdot 3 (same as previous, just reordered).
The smallest is 60. Answer: 60. Notice the strategy: to minimise n, assign the largest exponents to the smallest primes.
Variant 3: "Common divisors of a and b"
How many positive common divisors do 288 and 432 share?
The common divisors of a and b are exactly the divisors of \gcd(a, b). So this question becomes: "how many divisors does \gcd(288, 432) have?"
288 = 2^5 \cdot 3^2, 432 = 2^4 \cdot 3^3. So \gcd = 2^4 \cdot 3^2 = 144. Number of divisors: (4+1)(2+1) = 15.
Two exponent-level steps: compute the minimum exponent at each prime (for the GCD), then apply the divisor-count formula. No listing.
Variant 4: "Count perfect-square divisors"
How many positive divisors of 720 are perfect squares?
720 = 2^4 \cdot 3^2 \cdot 5. A divisor 2^{b_1} \cdot 3^{b_2} \cdot 5^{b_3} is a perfect square iff every exponent is even. So b_1 \in \{0, 2, 4\} (3 choices), b_2 \in \{0, 2\} (2 choices), b_3 \in \{0\} (1 choice — 5's exponent in 720 is 1, and the only even value \leq 1 is 0). Total: 3 \cdot 2 \cdot 1 = 6 square divisors.
The constraint "even exponent" cuts each factor of (a_i + 1) down to \lfloor a_i / 2 \rfloor + 1. Once you notice that, the problem is a one-liner.
Why this specific instinct matters
The divisor function \tau(n) is the gateway to an entire area of number theory called multiplicative functions. Once you have \tau, the same exponent-picking argument gives you:
- \sigma(n), the sum of divisors: for each prime, sum 1 + p + p^2 + \cdots + p^{a_i}, then multiply.
- \varphi(n), Euler's totient (count of integers \leq n coprime to n): n \prod_i (1 - 1/p_i).
- The number of square-free divisors: 2^r where r is the number of distinct prime factors.
- The number of coprime pairs in various ranges.
All of these share the same DNA — compute something at each prime, then multiply. If you train yourself to spot the pattern "question is about divisors / coprimality" → "translate to exponent counting," you have unlocked the whole family of problems.
Indian school and Olympiad context
This recognition appears constantly in:
- NTSE and CBSE. "Find the smallest number with [exactly k / at least k] divisors" appears year after year.
- JEE Main/Advanced. Often in combination with sets: "how many triples (a, b, c) satisfy \operatorname{lcm}(a, b, c) = \ldots?"
- Regional Math Olympiad. Problems about the divisor function's extremal behaviour — smallest n with \tau(n) = 100, largest n with \tau(n) = \tau(n-1), etc.
None of these yield to listing divisors. All yield to choosing exponents.
A gotcha to watch for
The formula \tau(n) = \prod (a_i + 1) counts positive divisors. If a problem asks for "divisors including negatives," double it (each positive d pairs with -d). If a problem asks for "proper divisors" (excluding n itself), subtract 1. If "non-trivial" (excluding 1 and n), subtract 2 — except when n = 1, in which case be careful.
These small adjustments happen after the exponent count. The hard part is always the exponent count itself.
One-line takeaway
"How many divisors?" is a trick question. The real question is always "how many ways to choose an exponent at each prime?" Factor n, read off the exponents, add 1 to each, multiply. Any follow-up ("divisors greater than / perfect-square divisors / common divisors") is a variation on the same theme — a constraint on the choices.
Related: Number Theory Basics · Factor Into Primes First · Smallest n Divisible by — Think LCM · Divisibility Tree Explorer · Sieve of Eratosthenes Animation