Most people want to make the right decision. This is the wrong goal. The right decision — the one that produces the best outcome — is only knowable in retrospect. At the time of decision, it is not available. What is available is the best decision you can make given what you know, and the discipline to make it through a process you can stand behind.
A defensible decision is one made with accurate information honestly assessed, through a clear reasoning process, with explicit acknowledgement of the uncertainties involved. A defensible decision may still produce a bad outcome — the world is partly random. But it will not produce a bad outcome because of a broken process. And it will not produce regret about the process, even if it produces disappointment about the result.
The process I use: identify the decision clearly. What exactly is being decided? Not the general direction — the specific choice. State the options. At least two; sometimes more. For each option, identify the best realistic outcome and the worst realistic outcome. Assess the probability of each. Choose the option with the most favourable expected value, weighted for the magnitude of the downside.
Write the reasoning down. Before the outcome is known. This is the most important step, and the most often skipped. Writing the reasoning before the outcome removes the possibility of outcome bias — of remembering your reasoning as better than it was because the outcome was good, or as worse than it was because the outcome was bad. The written reasoning is the actual record of the quality of your process.
The profound insight is that decision quality is a skill, like any other, and it can be developed. But it can only be developed if you have an accurate record of your reasoning — if you know what you actually thought at the time, not what you remember thinking in retrospect. Write it down. Review it. Notice where your reasoning was sound and where it was not. Improve the process. The outcomes will follow.