There are two different ways in which one can be unhappy about a decision one made, and they should be distinguished: you can be unhappy about the decision and think that, all things considered, you made the best choice possible; or you can be unhappy about it and think to yourself, ‘I should have known better, I should have made a different choice’. Now this latter form of unhappy decisions are of greater importance, for those are the one we feel worse about but have more to gain from looking at. I call these, for now, regretful decisions. The former type of decisions, of which we may feel sorry but all things consider we do not regret, are not very interesting – for it is, I think, quite a consensus among people who think about decision theory that the so-called ‘test of the result’ is an unimportant one. If you made the right decision and things turn out wrong, you shouldn’t think about it too much – there’s nothing you could do and nothing you could have done. If you took a relatively safe gamble (for in real life there are no sure gambles) and it didn’t work out as it was most likely to – there’s nothing to regret. This is the first psychological problem we must overcome – we have a tendency to evaluate actions according to their result. As a moral theory this idea has been powerful for a long time, and though it is often scolded by high-brow academics it is still very strong. As a decision-making theory, it’s downright silly. Nevertheless, we don’t only beat ourselves over right decisions that turned out not to lead to good results, but we judge and punish others – public servants for example – using result-based evaluation. This is simply a bad move on our part.
A second mistake is to confuse decisions of the two kinds. This is slightly different than the previous point - because we put so much weight on the results we tend to be mistaken in our own appreciation of the decision. Not only do we regret what we shouldn’t but we may also make wrong inferences about our decision-making mechanism and make bad changes. We may, coming to a similar situation in the future, make a different decision. This is just damaging ourselves twice and is very unfortunate when it happens. There is a subtlety involved – for my account is based on the idea that a decision should be evaluated on the basis of its plausibly expected results and not on its actual results in a particular occasion. Somebody may say, and rightfully so, that any such evaluation of expected results is necessarily uncertain while actual results from past experience are certain. I am the last one to oppose learning from one’s experience and the fact that something happens in the past certainly points its possibility – but surely not to its certainty. The uncertainty of the future remains even in face of the certainty of the past and it’s only people who treat time like a forth-dimensional continuous and uniform line that obscure this fact. In any case, this is a crucial fact for decision-making theory – every time you make a decision, the future is equally uncertain.
Finally we turn to regretful decisions – which I believe are the most interesting case. They present a problem – on the one hand, they present an opportunity to learn. If I made the wrong decision though I had all the necessary information, I would want to know why I did it. That’s even more pertinent if, as is usually the case, I tend to make this mistake often. If I can look at all my mistakes and look for systematic errors, I may be able to avoid them in the future – by avoiding particular types of decisions (under pressure or in the face of somebody who is doing their best to make you feel guilty) or just try hard to not to the mistakes I know that I have a tendency of doing. On the other hand, the exercise of going through my own mistakes is bound to take its toll on me. Indeed, I can’t think of many things that are more like beating myself over the head repeatedly with a flat object or just slapping my hand against my ever growing forehead repeatedly, possibly exclaiming “D’oh!”. This isn’t an easy trade-off, and despite my excessive irrational aesthetic strive for perfection I admit that it may be that it is often preferable to not dwell on mistakes and move on. The chance that I can actually change my tendency to make the same mistake again are not worth the self-inflicted pain of the interrogation.
Indeed, presented with this kind of calculus it may seem that we would need to make a decision whether or not to look into our decision, and of course this decision may be a mistake and we would enter a recursive process in which we would need to decide if we want to review the way we review our decision. Of course, this is not cyclic or problematic – I’m assuming that we would easily stop at the third or fourth order, deciding that it’s not worth our while to keep deciding whether or not we should review or decision making and stop at that. Indeed it seems self-defeating to think that we would go on and on forever, because at every stage we consider the process as costly and the returns are diminishing. Nevertheless, I don’t really like this second order calculus with its implications; regret is, after all, an emotion that we cannot completely control, and many people really would like to avoid it. Perhaps then, just on this basis, perhaps we want to examine regretful decisions – so that the next time we feel regert on our decision we won’t regret not examining them.