What makes a good decision good and a bad one bad? The quick answer is, “It depends.” It depends on how the decision turned out. If you go to a party and have a bad time, it was a bad decision to go, but if you go and meet the love of your life, it was a good decision. “It depends” may be a common way to think about decision quality, but is it the right way?

The practice of judging decisions by their results is sometimes called “resulting” in the world of professional poker, and (hint) it’s not considered wise. You can find a discussion of resulting in the book Thinking In Bets by poker champion Annie Duke — that’s where I first heard the term.

The problem with resulting is that results often contain randomness. Resulting, as an approach to evaluating decisions, fails to account for that randomness.

When we make any decision in the face of uncertainty, there’s a chance of a bad outcome, no matter how much skill and intelligence went into the decision process. If we equate decision quality with result quality, treating them as one and the same, we’re implying that decision quality is itself a matter of chance, which doesn’t make much sense.

There’s got to be a way of separating decision quality from result quality, so that we could still consider a decision “good,” even if it yielded a bad result due to the inherent randomness of the situation. Why not say that a good decision is one that followed a good strategy, regardless of the outcome the decision happened to have? This idea of focusing on a decision’s underlying strategy is an “antidote” to resulting.

Say you’re betting on a coin toss, and you know the coin is weighted so the chance of heads is 60%. It’s a good decision to bet on heads, because heads is more likely than tails. But this good decision can still have an unfortunate outcome: the coin might land on tails the first time it’s flipped, and you’ll lose the bet. Depending on what’s at stake, you might experience this loss as crushing and traumatic. If the sting of defeat makes you abandon your overall strategy of betting on heads, then you’ve fallen prey to “resulting.” And you’re not guaranteed to be cured of resulting on your second try. What if you change strategies and bet on tails next time? You might well be rewarded — there’s always a 40% chance of tails. If you get lucky with tails on the second bet, this confounding reward might further solidify your distrust of the always-heads approach. But such a growing skepticism of heads and an increasing attachment to tails is only setting you up to lose in the long run.

I don’t think you’re likely to give up on heads though — not if you’re certain about the coin’s 60% bias. More likely, after losing the first bet, you’d recognize that betting on heads was still the best strategy you could have applied. You’d bet on heads again, and again, and after some time, you’d reap the reward.

But what makes this scenario so transparent? What makes this coin toss easy to think about, in comparison to the scenarios we encounter in everyday life? What makes it possible for you to avoid the trap of resulting here, while that’s so hard to do elsewhere? Why are you not likely to be wracked with self-doubt and regret after losing your first bet on heads here, whereas there are major life decisions that you might still regret? What allows you to maintain your confidence and self-respect as you continue to bet on heads in this situation, whereas similar confidence can be so precarious to maintain in the face of life’s many setbacks and bad outcomes?

There are two things working in your favor as you process the pain of your failed bet on heads: first, your betting strategy is clear and rational, and second, your bet is repeatable. Anyone who loses any kind of bet might find themselves wondering, “Why did I lose? Did I do something stupid?” But in this scenario you can remind yourself that you followed a sound strategy. You used your knowledge of the coin’s bias, and that’s all the knowledge there was to use. Although the “goodness” of your strategy isn’t revealed by the first or even the second outcome it happened to have, you can still experience that goodness by flipping the coin several more times. If you keep betting on heads, you’ll see that this strategy works out better than any other one you could come up with. Repeated many times, this strategy will eventually produce evidence of its goodness.

Wouldn’t it be nice if everyday decision scenarios were just as easy to reason about as this coin toss? Why aren’t they? When we make an everyday decision, why is it often hard to judge whether we made a good one or a bad one without looking at the result? That’s because when we make an everyday decision, 1) our strategy might not be clear, and 2) the decision might not be repeatable. If we’re questioning the quality of a past decision, we can’t always go back and identify a coherent strategy that we might have used, and we can’t always “try out” the same kind of decision again. What are we left with as we try to assess the decision? The result stares us in the face. The result is all we have to go by.

What if you decide to accept a party invitation, and you have a bad time? You come home and regret your decision, because now you’re behind on chores, and you’ve also missed out on quiet time. Was it still a good decision to go to the party? In a wild dream, perhaps you’d be able to say, “I was just applying my proven strategy for choosing when to accept party invitations. The strategy didn’t give the result I wanted this time, but I’ve been applying that strategy consistently, and I’m confident that it is a good strategy overall.”

The problem is, you didn’t apply a “proven strategy” when you accepted this particular invitation, right? Be honest — how did you make the decision? Perhaps you wavered for a while, thought about this and that, looked at your schedule, felt some hope, felt some doubt, decided to turn it down, waited a while longer, found out that a certain guest was going, and another had declined. There was flux elsewhere on your calendar, and then you changed your mind and accepted this invitation at the last minute. The decision feels totally unique — it’s a “one off” — it’s “yours” — it’s a product of your inexplicable, irreducible self, responding to your own complex life circumstance in a somewhat unpredictable way. You don’t really know how you decided, but you feel like the whimsical decision “came from you,” not from any specific policy that you were cooly enacting.

Surely, there exists a strategy that would have generated the same decision you made. Such a strategy might be, “If you’re looking for love, go to as many parties as you can.” But just because such a strategy exists and aligns with your decision doesn’t mean it’s the strategy you actually used at the time. The “looking for love” strategy can be cited, after the fact, to explain why you should have accepted the invitation, but it’s not necessarily why you did. Even if you were aware of this strategy, and were influenced by it, you might have had opposing strategies in mind as well, like, “If you need rest, stay home.” In the end, your decision isn’t attributable to any single strategy. Rather, it’s the outcome of a clash between multiple opposing strategies battling it out in your mind, along with emotions and scheduling considerations and whatnot, finally crowning an uncertain and seemingly random victor.

To judge whether your decision, despite its bad outcome, is still a good kind of decision overall, can you repeat it? There will be more parties invitations in your life, but when? Perhaps it’s a slow season, socially, and you have other competing obligations. It could be months till the next invite comes along that you can accept. Practically, you’d find it hard to accept a few dozen party invitations in rapid succession, to test out your strategy, in the same way you could flip a coin several dozen times and look for a trend. Each party is unique — different people, different location, different theme — so it’s not like one party is “replaceable” by the next. And are you really keeping records of which invitations you accept and decline and how well each decision turns out?

Another challenge you face in assessing your decision is that “Hindsight is 20/20.” If you didn’t have a good time at the party, you might think you should have known it was going to be a waste: “I was turned off by that flowery cursive font on the invitation card, and I didn’t like the formal dress requirement. I should have paid more attention to those clues and known this was going to be a stuffy affair.” But you didn’t pay special attention to those clues because you didn’t yet know you’d have a bad time.

After you learn the outcome of a decision, you can always identify ways that you could have predicted that outcome. And if you could have predicted the outcome, you could have used your prediction to make a different decision that would have given a better outcome. But that doesn’t mean the outcome was predictable without the knowledge of what it was going to be. That’s because there may have been elements in the situation that could have predicted the opposite outcome too. Some clues pointed to a bad outcome, and others clues pointed to a good one. There was a basis for pessimism and a co-existing basis for optimism. You didn’t have enough information, at the time, to know which of the competing clues were the most important.

If the party had turned out to be wonderful and indeed you met the love of your life there, you could say, “I should have known it was going to be amazing. I shouldn’t have worried so much. I should have remembered that the hosts are good friends who always invite interesting guests, and I should have trusted my inclination to go.” But you didn’t pay special attention to those clues because you didn’t yet know you’d have a good time!

If we’re going to avoid “resulting,” we need to differentiate decision quality from result quality. We can do that by looking at the strategies that produce our decisions, and accepting that a decision was “good” if it followed a good strategy. But we’ve seen why that’s much harder to do in everyday life than it is in a coin toss. In everyday life, we can’t always identify a clear strategy behind the decisions we’ve made. And even when we can identify a strategy that would have generated the same decision as ours, we can’t always repeat that strategy many times in a row to gain evidence for that strategy’s goodness. To make things even harder, once we know an outcome, we might be tempted to look back and see how a different strategy might have led to a better outcome.

Our everyday choices are “messy” and our decision processes can be indecipherably complex, so we feel ownership over our decisions, as though they come from the deepest, inexplicable depths of our inner self. And so, when a decision leads to a bad outcome, we feel a challenge to the self. This dynamic keeps us trapped in a place where we’re often regretting and second-guessing our decisions. Randomness guarantees that we’ll always experience some unfortunate outcomes, no matter what we do, so we’ll always have decisions to regret, no matter what we do.

How can we fix this? It would be unrealistic to recommend that we should fix it by making all our everyday decisions according to clear strategies. I don’t think we are, or should ever aspire to become, embodied algorithms. Life is always going to be messy, and we’re always going to have to make messy decisions. But when we can use clearer strategies and name them beforehand, perhaps we should.

It would also be unrealistic to recommend that we should test our decision strategies by applying the same strategy repeatedly and gathering evidence for its effectiveness. Some decisions are “one of a kind,” and life doesn’t always give us to the opportunity to make similar ones again. But when we can repeat a decision strategy and look for trends in the outcomes it creates, perhaps we should.

When a decision isn’t repeatable and there isn’t any coherent strategy that we can locate behind it, we can still look for a meta-strategy that we might have applied implicitly. No matter how a particular decision came about, for example, we probably applied the meta-strategy that, “It’s better to make a decision than to indefinitely avoid making one.” And if that’s a good meta-strategy, we can give ourselves some credit for applying it. We can give ourselves some credit for deciding — even if it took a while — rather than deferring forever.

Finally, if we remain aware of the perils of resulting, we can stop ourselves from doing it when we notice we’re doing it. For that to be possible, we need to understand what resulting is, and why it’s so tempting, and hopefully now we do. ■

Comments ༄