Ponder This Challenge - February 2026 - Blot-avoiding backgammon strategy
- Ponder This
This riddle was suggested by Latchezar Christov - thanks Latchezar!
We will call two triangles "sibling pairs" if they are two non-identical triangles with integer side lengths respectively and , where are the only side of each triangle with a different length from the other triangle, since there are two common side lengths **and** both triangles must have a common nonzero area.
An example is the pair of triangles with side lengths and . The area of both is 12. Another example is and with area .
For some and , there may not be any sibling pair at all (e.g., and ) and for some, there may be more than one. For example, if and , three sibling pairs exist with the following side lengths:
Your goal: Find two integers such that and there are **exactly** 50 sibling pairs with common side lengths and .
**A bonus** "*" will be given for solving the above problem with the additional constraint that at least two of those 50 sibling pairs must have **integer** areas. The areas of the remaining pairs may be irrational.
September 2024 Solution:
A solution to the main problem is
A solution to the bonus is
The area of a triangle with sides a,b,c is given by Heron's formula: denote , and the area is . Thus, if we want to find such that the area of the triangles and is equal, it sufficies to satisfy the equation . Assuming this simplifies to , which is related to a classical problem in number theory - the representation of an integer N as the sum of two squares.
Given N, finding such representation can be hard unless one has the factorization of N, and then one case use a recursive algorithm for finding all such representations thanks to the Brahmagupta–Fibonacci identity
Which shows how to obtain a representation as sums of squares of a number from such representations of .
We are looking for solutions to to the equation where . However, not every which satisfy the equation lead to a triangle, since the area might turn out to be zero or negative: we need to have , assuming . We proceed as follows, given some :
To solve the problem, one iterates on various values of . It is best to look at the first few primes congruent to 1 modulo 4, since product of such primes is guaranteed to be representable as sum of squares. The relatively small solution can be found using (where 5,13,17 are the first primes congruent to 1 modulo 4). The even smaller solution is obtained by considering two more primes: .
Finding solution to the basic problem is extremely quick given efficient implementations for steps 1-3, but finding a solution to the bonus can still be difficult.