We all know that there are two extremes in any action we do or the decisions we take. Let's say one has no confidence and other has overconfidence, one is slow other is fast and I observed this trend followed in fields in finance(risk-taking), Machine Learning (finding the optimum learning rate). While learning any new task too there ought to be the right amount of difficulty to have a good learning curve. People generally in finance say we should be risk-averse but without having an emergency fund they won't take risks. We can only find these sweet spots by experimenting and learning from our mistakes.