Complexity earns its place only when simplicity cannot solve the problem.

Complexity is often adopted because it looks like rigor. But every additional part has to be understood, maintained, explained, and eventually repaired, and those costs are paid continuously while the benefit may be occasional. The question is not whether the complicated version is better in theory but whether the simple version actually fails. Most of the time it has not been tried.

When you are about to add something, ask what the simplest version would fail at specifically. If you cannot name the failure, use the simple version. Add complexity only after simplicity has demonstrably broken.