'There is an important point to note here. Poor outcomes are different from poor decisions. The best decision possible given the knowledge available at the time the decision is made can still turn out unhappily. A patient may die despite receiving the correct treatment. A poor decision is one in which we regret the process we used. A person will consider a decision to be poor if the knowledge gained would lead to a different decision [next time]. Knowing simply that the outcome was unfavourable should not matter. Knowing what you failed to consider would matter.'
In medicine, we have a system called 'morbidity and mortality meetings' where cases in which things went wrong are reviewed to see if the team can learn from what happened. However, as Shiralkar points out, these are often like giving people feedback on their driving by listing the cars they hit, without checking out their vision. Lots of research shows outcome bias and hindsight bias have important clinical implications. 'Hindsight bias reduces our ability to learn from the past. It permits us to think we were better at making decisions that we really are, and can result in us being more confident about the accuracy of future decisions than we have a right to be. For example, cases presented in meetings may seem predictable because of hindsight bias. You may conclude you know this already and consequently fail to learn from the lessons illustrated by the case ... many errors appear transparent in hindsight. However, hindsight does not take into account the prevailing conditions at the time the decision was made. When we know the outcome, it significantly influences how we perceive past events.'
When things go wrong, this is important to remember if we are to learn from what has happened. For further fascinating reading, see this article: Hindsight bias, outcome knowledge and adaptive learning.