Whether a decision as such is good or bad is never determined by its actual consequences as such.
Decisions are made before their consequences are reälized (made actual). Instead, decisions are made in the face of possible consequences. There may be an ordering of these consequences in terms of plausibility, in which case that ordering should be incorporated into the making of the decision. Most theories even presume that levels of plausibility may be meaningfully quantified, in which case (ex hypothesi) these quantifications should be incorporated into the process. But even in a case where there were only one outcome possible, while the decision could (and should) be made in response to that unique possibility, it still were possibility of the consequence that informed the decision, and not actuality. (Inevitability is not actuality.)
When the reälized consequences of a decision are undesirable, many people will assert or believe that whoever made the choice (perhaps they themselves) should have done something different. Well, it might be that a bad outcome illustrates that a decision were poor, but that will only be true if the inappropriateness of the decision could have been seen without the illustration. For example, if someone failed to see a possibility as such, then its reälization will show the possibility, but there had to have been some failure of reasoning for a possibility to have ever been deemed impossible. On the other hand, if someone deemed something to be highly unlikely, yet it occurred anyway, that doesn't prove that it were more likely than he or she had thought — in a world with an enormous number of events, many highly unlikely things happen. If an event were highly unlikely but its consequences were so dire that they should have been factored into the decision, and yet were not, the reälization of the event might bring that to one's attention; but, again, that could have been seen without the event actually occurring. The decision was good or bad before its consequences were reälized.
A painter whose canvas is improved by the hand of another is not a better painter for this, and one whose work is slashed by a madman (other than perhaps himself) is not a worse painter for that. Likewise, choosing well is simply not the same thing as being lucky in one's choice, and choosing badly not the same as being unlucky.
Sometimes people say that this-or-that
should have been chosen simply as an expression of the wish that more information had been available; in other cases, they are really declaring a change in future policy based upon experience and its new information. In either case, the form of expression is misleading.
Some readers may be thinking that what I'm saying here is obvious (and some of these may have abandoned reading this entry). But people fail to take reasonable risks because they will or fear that they will be thought fools should they be unlucky; some have responded to me as if I were being absurd when I've referred to something as
a good idea that didn't work; our culture treats people who attempt heinous acts but fail at them as somehow less wicked than those who succeed at them; and I was drawn to thinking about this matter to-day in considering the debate between those who defend a consequentialist ethics and those who defend a deöntological ethics, and the amount of confusion on this issue of the rôle of consequences in decision-making (especially on the side of the self-identified consequentialists) that underlies that debate.