AMC 10 · 2016 · #5

Grade 6 number-theory
logical-deductionif-then-reasoningprime-numbersparity contrapositivecasework ↑ Prerequisites: logical-deductionprime-numbers
📏 Medium solution 💡 2 insights
Problem
A claim says every number of one kind can be split a certain way. Find what would disprove it.

Pick an answer.

(A)
an odd integer greater than } 2 \text{ that can be written as the sum of two prime numbers
(B)
an odd integer greater than } 2 \text{ that cannot be written as the sum of two prime numbers
(C)
an even integer greater than } 2 \text{ that can be written as the sum of two numbers that are not prime
(D)
an even integer greater than } 2 \text{ that can be written as the sum of two prime numbers
(E)
an even integer greater than } 2 \text{ that cannot be written as the sum of two prime numbers
How to solve
Strategy Change Focus / Count the Complement

The conjecture is a claim about every even number greater than 2, so breaking it is not a matter of finding a pattern — it is a matter of finding one number where the claim fails. Switching focus from "all cases work" to "one case fails" produces an exact template with two required halves, and each of the five descriptions can then be held against that template. Concrete test numbers finish the job: a description that is already satisfied by some number alive today cannot be a counterexample, because the problem says no counterexample has ever been found.

1STEP 1

Write the claim as if-then

The claim is one if-then statement.

for every integer n: (n even and n > 2) → n = p + q for some primes p, q
2STEP 2

See what breaks a for-every claim

A counterexample needs both halves.

counterexample n: (n even and n > 2) and n ≠ p + q for all primes p, q
3STEP 3

Odd numbers cannot testify

A number of the wrong kind cannot testify.

9 = 2 + 7 (both prime); 11 = 2+9 = 3+8 = 5+6 = 7+4 (never two primes)
4STEP 4

A bad split is not a failure

A single bad split is not a failure.

8 = 4 + 4 (neither prime), 8 = 3 + 5 (both prime)
5STEP 5

An example is not a counterexample

A working example is not a counterexample.

2016 = 13 + 2003 (both prime) → conjecture holds for 2016
6STEP 6

Keep the one that fails

Only one candidate breaks it, choice (E).

even, n > 2 (condition met) + no prime pair sums to n (promise broken)
Answer
an even integer greater than 2 that cannot be written as the sum of two prime numbers
Check (E) against the conjecture directly: if such a number existed, the sentence "every even integer greater than 2 is a sum of two primes" would be false, because that number is even, is greater than 2, and is not such a sum. Now check the other four from the opposite side. Each of them is already realized by a number available today — 9 for (A), 11 for (B), 8 for (C), 2016 for (D) — and the problem states that no counterexample has ever been found. So none of those four descriptions can describe a counterexample, and only (E) survives.
💡Key takeaway

To break a claim that says "every", you need one case that fits the claim's condition and still fails its promise — nothing weaker counts.

  • Write the claim as if-then
  • See what breaks a for-every claim
  • Odd numbers cannot testify
  • A bad split is not a failure
  • An example is not a counterexample
  • Keep the one that fails