Sieve of Eratosthenes vs Atkin

I’ve seen many solutions for problems (namely CHEFHACK and PRIME1) that deal with prime numbers, and the algorithm most people choose to generate these prime numbers is the Sieve of Eratosthenes.

However, from my understanding, the Sieve of Atkin is much faster than the Sieve of Eratosthenes. I even had first hand experience with both algorithms with CHEFHACK and determined that the Sieve of Atkin was significantly faster.

Why is it that people tend to choose Eratosthenes over Atkin?

Eratothenes is easy to code and understand

And the fact that most algorithmic contest problems are constructed to accept solutions using Eratosthenes.