One Strategy for a Whole Year?

Many traders change strategies after a few losses. What if the real solution were to master one strategy and collect enough data before changing anything? How long do you test a strategy before deciding whether it works?