Specifically, Big-O describes the relative performance of an algorithm as the input grows. That's why constants are discarded, and that's why O(n^2) is probably

Specifically, Big-O describes the relative performance of an algorithm as the input grows. That's why constants are discarded, and that's why O(n^2) is probably fine if you only ever have very small data sets.