How to Decide: Simple Tools for Making Better Choices
von Annie Duke
Because there are only two things that determine how your life turns out: luck and the quality of your decisions. You have control over only one of those two things.
Any decision is, in essence, a prediction about the future. When you’re making a decision, your objective is to choose the option that gains you the most ground in achieving your goals, taking into account how much you’re willing to risk. (Or sometimes, if there aren’t any good options, your objective is to choose the option that will cause you to lose the least amount of ground.)
Because there are so many possible futures, making the best decision depends on your ability to accurately imagine what the world might look like if you were to choose any of the options you’re considering.
Determining whether a decision is good or bad means examining the quality of the beliefs informing the decision, the available options, and how the future might turn out given any choice you make.
If your decision process becomes better than it is now—improving the accuracy of your knowledge and beliefs, improving how you compare available options, and improving your ability to forecast the futures that might result from those options—that is worth pursuing.
You can spot this phenomenon across all sorts of domains. You buy a stock. It quadruples in price. It feels like a great decision. You buy a stock. It goes to zero. It feels like a terrible decision.
In every domain, the outcome tail is wagging the decision dog. There’s a name for this: Resulting.
Psychologists call this “outcome bias,” but I prefer the more intuitive term “resulting.”)
RESULTING A mental shortcut in which we use the quality of an outcome to figure out the quality of a decision.
A necessary part of becoming a better decision-maker is learning from experience. Experience contains the lessons for improving future decisions. Resulting causes you to learn the wrong lessons.
This is, of course, not limited to investment decisions. You quit your job to join a promising start-up because it offers you equity. It ends up as the next Google. Great decision! You quit your job to join a promising start-up because it offers you equity. It fails after a year. You end up out of work for six months and you run through your savings. Terrible decision!
The point of having you write those down is that most people don’t actually think much about their best and worst decisions. They usually start by thinking of their best and worst outcomes and work backward from there. That’s due to resulting.
Being a better decision-maker means being a better predictor of the set of possible futures.
When you overfit decision quality to outcome quality, you risk repeating decision errors that, thanks to luck, preceded a good outcome. You may also avoid repeating good decisions that, because of luck, didn’t work out. Resulting has the biggest effect on learning when outcome and decision quality are misaligned.
Other ways the decision could have turned out: Even without knowing much about the movie business, when a movie is just a concept, as Star Wars was when Lucas pitched it, a lot of things could happen. His concept could sound great but look terrible when executed, $10 million later. None of the stars were big names. If Lucas had cast different actors, the movie might have flopped. The mass audience could have decided it wasn’t interested in sci-fi films. A recession could have hit just as the movie was being released that kept people from going to the movies.
It seemed like a crazy project to them but the studio head told Lucas, “I don’t understand this, but I loved American Graffiti, and whatever you do is okay with me.”
The point is, it’s difficult to reach a conclusion about decision quality from one result. That one result shouldn’t count as much as a greater quantity of data (on all the decisions the studio executives made and their overall results) or data of higher quality (on what the decision looked like as it was presented to the studios).
Either way, your friends are saying, “I knew it!” It goes without saying that you can’t know you’ll hate it and know you’ll love it at the exact same time. But that’s how we all feel. So what gives? What gives is hindsight bias.
But there are obvious cues that signal hindsight bias, such as “I can’t believe I didn’t see that coming,” or “I knew it,” or “I told you so,” or “I should have known it.” Training yourself to listen for those mental and verbal cues is a good way to hone your hindsight bias spotting skills.
Memory creep is the reconstruction of your memory of what you knew that hindsight bias creates. MEMORY CREEP When what you know after the fact creeps into your memory of what you knew before the fact.
And here’s the thing: If you misremember the past, you are going to learn useless lessons from your experience. That can mess you up in two ways: You’re not going to remember what you knew at the time of the decision. That makes it hard for you to judge whether a decision was good or bad. To assess the quality of a decision and learn from your experience, you need to evaluate your state of mind honestly and recall what was knowable or not knowable as accurately as possible. Hindsight bias makes you feel like the outcome was much more predictable than it was. This can cause you to repeat some low-quality decisions and to stop making some high-quality decisions.
Using a Knowledge Tracker reduces hindsight bias by clarifying what you did and didn’t know at the time of the decision. Detailing what you knew and when you knew it helps prevent stuff that revealed itself after the fact from reflexively creeping into the before-the-fact box.
As you were using the Knowledge Tracker, it may have occurred to you that it would be a good idea to journal the “stuff you knew before the decision” while you are in the process of making the decision. It can be hard to accurately recall what you knew before the fact once you already know the outcome. Journaling gives you something concrete to refer back to.
went ten pages deep on Google and couldn’t find any critique specifically about Michigan, Pennsylvania, and Wisconsin from before the election. Although there is an abundance of opinion pieces critiquing other aspects of Clinton’s campaign strategy, none of them anticipated this particular problem.
WHY WAS DONALD TRUMP CAMPAIGNING IN JOHNSTOWN, PENNSYLVANIA? (WashingtonPost.com, October 22, 2016) WHY IS DONALD TRUMP IN MICHIGAN AND WISCONSIN? (NewYorker.com, October 31, 2016)
There are many possible futures, but only one past.
If there are ten pros and five cons, does that mean you should go with the decision? It is impossible to say without information about the size of the payoffs, because without that you can’t figure out if the upside potential outweighs the downside.
This way of thinking, that there is only “right” and “wrong” and nothing in between, is one of the biggest obstacles to good decision-making. Because good decision-making requires a willingness to guess.
Here’s a secret: All guesses are educated guesses because there is almost no estimate you could make about which you literally know nothing.
Just like decision-making, archery is not all or nothing, where you get points only for hitting the bull’s-eye and everything else is a miss. An archer gets points for hitting the target at all. The value of guessing isn’t in whether the guess is “right” or “wrong.” Your guesses are like the archer’s arrows. If you were omniscient and your guesses were always exactly right, you’d score all bull’s-eyes. When you make an educated guess, you’re aiming at the bull’s-eye and, though it’s likely you’ll miss the exact answer, like the archer you will still score points for getting in the vicinity.
Your choice is always an estimate of the likelihood of different outcomes unfolding.
Because a lot of the stuff you don’t know lives in other people’s heads, getting feedback from other people about the things you believe and the decisions you make will be one of your best tools for extracting knowledge from the world.
When President Kennedy approved the CIA plan to overthrow Fidel Castro (known as the Bay of Pigs invasion), he asked his military advisers for their opinions about whether the attempt would succeed. The Joint Chiefs of Staff told Kennedy the CIA’s plan had a “fair chance” of success (which the writer of the assessment considered to be 25%). Because Kennedy thought “fair chance” meant something much higher, he approved it. The plan was a failure, which looked clumsy and amateurish, and embarrassed the United States at a key moment in the Cold War.
A convenient way to express where you are on the continuum from no information and perfect information is to offer, along with your exact (bull’s-eye) estimate, a range around that estimate. That range communicates the size of your target area by giving the lowest reasonable value you think the answer could be (the lower bound) and the highest reasonable value you think it could be (the upper bound).
Having a wide range is not a bad thing. Rather, it is a way to reflect as accurately as possible how much educated there is in your educated guess. A wider range that is true to what you do and don’t know is more helpful than a narrower range that oversells your certainty.
Wharton professor Abraham Wyner suggests that a good way to get to a reasonable upper and lower bound is to ask yourself, “Would I be pretty shocked if the answer fell outside this range?” If you use that as your standard, your range will naturally reflect how much educated there is in your educated guess.
As an outside observer, you can see this clearly when you are in the “friend” position. But your vision becomes muddied when you are on the inside and it’s your problem. What you can see so clearly in others is hard to see in yourself. That’s why almost everybody feels like they are better at solving other people’s problems than their own.
More than 90% of professors rate themselves as better-than-average teachers. About 90% of Americans rate their driving ability as better than average. Only 1% of students think their social skills are below average.
The person you’re most likely to mislead is yourself. And you don’t know you’re doing it, because you’re living in the inside view.
Base rates: An easy way to get the outside view One way to get to the outside view is to make it a habit, as part of your decision process, to ask yourself what is true of the world in general, independent of anyone’s point of view. A helpful way to get an idea of what’s true of the world in general is to find out if there is information available about the likelihood of different outcomes in situations similar to yours. That information is called a base rate.
That’s why, in the first exercise, most people answer that the things they are thinking to themselves are different from the things they would say out loud to their friend who always has a bad dating story to share. You are trying not to hurt your friend’s feelings. You are trying to be kind. But in doing so, you’re denying them valuable input that could improve the quality of their future dating decisions. In being kind to your friend in the now, you are being unkind to the versions of your friend in the future who will have to make new dating decisions.
Here are two tactics you might try in getting to the outside view: (1) Ask yourself, if a coworker or friend or family member were to have this problem, how would you view their problem? How might your perspective differ from theirs? What advice might you give them? What kind of solutions would you offer? (2) Ask yourself if there are any relevant base rates or information you could find about what’s true of people in your situation in general.
Pros and cons lists amplify the inside view.
You should be eager to hear people disagree with you and motivate them to do so.
Here’s how long the average person spends per week on these decisions: What do you want to eat?—150 minutes a week. What do you want to watch on Netflix?—50 minutes a week. What do you want to wear?—90 to 115 minutes a week. This means that if you’re like most people, you’re spending a lot of time in analysis paralysis. The time the average person spends deciding what to eat, watch, and wear adds up to 250 to 275 hours per year. That’s a lot of time spent on decisions that intuitively feel like they are inconsequential.
It may seem that spending an extra minute of your time here and there on these routine decisions isn’t that big a deal, but that’s because it’s a death by a thousand cuts. These tiny expenditures mount up over time until you have spent seven workweeks a year deciding what to eat, watch, and wear.
THE TIME-ACCURACY TRADE-OFF Increasing accuracy costs time. Saving time costs accuracy.
The key to achieving the right time-accuracy balance is figuring out what the penalty is for making a lower-quality decision than you would have if you had taken more time. How much leeway is there to sacrifice accuracy for speed? The smaller the penalty, the faster you can go. The bigger the penalty, the more time you should take on a decision. The smaller the impact of a poor outcome, the faster you can go. The bigger the impact, the more time you should take.
Even though many decisions won’t have a significant impact on your long-term happiness, there is still a short-term cost of a bad result: regret. Regret (or fear of regret) can make you indecisive about nearly any choice.
Options repeat for decisions where you will get another crack at the same choice, which is especially useful if the choice comes up again quickly. You might really dislike the dish you order in a restaurant, but you will get another chance to pick something to eat in just a few hours. And that helps take the sting out of any short-term regret. Choosing college classes is a repeating option. Choosing who to go on a first date with is a repeating option. Choosing driving routes is a repeating option. Choosing a movie to see is a repeating option.
FREEROLL A situation where there is an asymmetry between the upside and downside because the potential losses are insignificant.
The concept of a freeroll is a useful mental model for spotting opportunities you can decide to seize quickly. The key feature of a freeroll is limited downside, meaning there isn’t much to lose (but there might be a lot to gain). The usual penalty for speeding up—the possibility of a greater likelihood of a worse outcome—doesn’t apply when you are in freeroll territory.
You can identify decisions with limited downside by asking yourself one or both of the following questions: What’s the worst that can happen? If the outcome doesn’t go my way, am I worse off than I was before I made the decision?
The bigger the asymmetry between the upside and downside, the more you have to gain when potential losses are limited, and the bigger the freeroll.
You may think that freerolls are too good to be realistically available. But once you’re on the lookout for them, you’ll find that freerolls are more abundant than you think.
The faster you engage, the less likely it is that the opportunity goes away. The faster you decide to seize the opportunity, the faster you get the chance at realizing the one-sided, upside potential of the decision.
Once you identify a freeroll, you don’t need to think too hard about whether to seize the opportunity, but you still want to take time with the execution of the decision. Go fast deciding whether to apply to a college that is a big stretch, but take time making sure the application is high quality. Go fast deciding whether to offer on your dream house, but take time making sure the offer is sound.
When you pass on such opportunities or let those small, temporary negatives slow you down, you’re magnifying the moment of rejection and ignoring the asymmetry working in your favor. You’re saving yourself from those short-lived feelings if the opportunity doesn’t pan out, but you’re costing yourself the chance for a meaningful, long-term boost to your well-being.
Warning: A free donut is not a freeroll When considering whether a decision has limited downside potential, it’s crucial to think about the cumulative effects of making the same decision repeatedly, rather than focusing on just the onetime, short-run potential harm.
When asking yourself “What’s the worst that could happen?” make sure you follow up by examining the effects of making the same type of decision repeatedly. That’s how you recognize that a free donut is not a freeroll.
That tells you that the closeness of the options is what’s slowing you down. You’d have no trouble choosing between options as far apart in their potential payoffs as a week in Paris versus a week spent among discarded fish parts. And that’s a clue as to why you can and should speed up these types of decisions.
When a decision is hard, that means it’s easy The very thing that slows you down—having multiple options that are very close in quality—is actually a signal that you can go fast, because this tells you that whichever option you choose, you can’t possibly be that wrong, since both options have similar upside and downside potential.
From that vantage point, the decision looks like a wolf, a dangerous, high-impact beast of nonrepeating options and lots of potential downside. Close calls might feel like the wolf is at your door. But this type of decision is really a sheep in wolf’s clothing.
That question allows you to think prospectively, understanding that what matters for decision quality is the potential of each of the options, not which of many possible outcomes happens to be the one that unfolds. That question allows you to see that you have two similarly great options to choose from, so whichever option you go with, it’s unlikely you’re making that big a mistake.
This unlocks a powerful decision-making principle: When a decision is hard, that means it’s easy. WHEN A DECISION IS HARD, THAT MEANS IT’S EASY When you’re weighing two options that are close, then the decision is actually easy, because whichever one you choose you can’t be that wrong since the difference between the two is so small.
That’s the paradox: more choice, more anxiety.
A useful tool you can use to break the gridlock is the Only-Option Test. If this were the only thing I could order on the menu . . . If this were the only show I could watch on Netflix tonight . . . If this were the only place I could go for vacation . . . If this were the only college I got accepted to . . . If this were the only house I could buy . . . If this were the only job I got offered . . . The Only-Option Test clears away the debris cluttering your decision. If you’d be happy if Paris were your only option, and you’d be happy if Rome were your only option, that reveals that if you just flip a coin, you’ll be happy whichever way the coin lands.
Menu Strategy This strategy for choosing what to order from a menu can be broadly applied to decision-making in general. For any decision, spend your time sorting the world into stuff you like and stuff you don’t like. After that, go fast. THE MENU STRATEGY Spend your time on the initial sorting. Save your time on the picking. The big gains that you get from your decision-making time are in the sorting: figuring out, given your values and your goals, what makes an option “good.” Sorting options is the heavy lifting of decision-making and that’s the place you will get the most value out of slowing down.
That’s why identifying low-impact decisions, especially ones that repeat, is so important. Those types of low-risk decisions give you the opportunity to experiment. Experimentation gets the world to tell you what works and what doesn’t work and helps you figure out your preferences, your likes and dislikes. And all that experimentation will make you better informed, paying off in more accurate sorting.
Quitting doesn’t deserve its nearly universal negative reputation. Quitting is a powerful tool for defraying opportunity cost and gathering intel, intel that will allow you to make higher-quality decisions about the things you decide to stick to. Whenever you choose to invest your limited resources in an option, you’re doing so with limited information. As your choice plays out, new information will reveal itself. And sometimes that information will tell you that the option you chose isn’t the best option for advancing you toward your goals.
Try stuff you can quit. Figure out what you like and what you don’t like. Figure out what works and what doesn’t work.
But being “quitty” allows you to make better choices about when to be gritty.
When you can do more than one thing at a time, you get many more opportunities to poke at the world, getting the input from multiple experiences.
Once you settle on a choice that’s good enough—regardless of how long you’ve taken, whether you’ve flipped a coin or conducted a lengthy decision process, or whether your options are indistinguishable or you have a clear favorite—part of a good decision process includes asking yourself a final question: “Is there information that I could find out that would change my mind?” You flipped a coin and it comes up “Paris.” Is there information you could find out that would make you switch your choice to Rome?
Allocate your decision time using the menu strategy. Spend time sorting, determining which options you like. Once you have options you like, save time picking.
Of course, you can rarely approach perfect information. If you’re wasting your time on illusory or infinitesimal gains in precision, you’re losing the chance to spend that time where the return is greater, or on better sorting, or on making more experimental choices that provide low-cost information for later decisions. That’s why many of the strategies laid out in this chapter are designed to steer you toward a more realistic approach to decisions known as satisficing (a term made from the combination of “satisfy” and “suffice”).
PROSPECTIVE HINDSIGHT Imagining yourself at some time in the future, having succeeded or failed at a goal, and looking back at how you arrived at that destination. Prospective hindsight enhances mental contrasting because looking back from your destination is a more effective way to plan the best route than looking ahead at where you are trying to go.
If you’re like most people, doing the premortem helped you identify some reasons for failure that you wouldn’t have thought of otherwise. Research suggests that when you combine mental time travel and mental contrasting, you can produce 30% more reasons for why something might fail. That’s an obvious upgrade for your crystal ball. A premortem boosts the clarity with which you can glimpse into the future. And the more complete your view of the future, the better your decision-making will be.
Like Ulysses, you can physically prevent yourself from making poor decisions. You can raise barriers, making it harder to execute on actions that will defeat your goals. When you raise barriers, you’re not physically preventing yourself from acting, like when you tether yourself to a mast. But you are increasing friction to make it more difficult to tamper with your plans. Raised barriers also provide you with a moment to stop and think before acting. You can lower barriers, reducing the friction to execute on actions that advance you toward success.
You want to get to work on time so you keep your alarm across the room, which raises a barrier to hitting the snooze button and falling back to sleep.
You decide to eat healthier and you know that late-night snacking is where you’ll go wrong. You can throw out all the junk food in your house. You could still have food delivered or go to the nearest drive-through, but getting rid of the junk food increases friction and raises the barrier to giving in to that impulse. You can also stock your house with healthy food or pack a lunch to bring to work with you. That reduces friction and makes it easier to stick to good choices.
This demand for context is especially relevant for team decisions. Team members should be encouraged to question decisions that have easily accepted rationales, looking for situations where exceptions have the potential to become the rule (and that rule undermines achieving the original goal). Second, you can make a precommitment to take those choices out of your hands by making a category decision. You already make category decisions all the time with dietary choices. If you’ve decided you’re a vegan, that’s a category decision—animal products are no longer an option when deciding what to eat. If you follow a Keto diet, simple carbs aren’t an option.
Evil’s grasp. A common practice of successful professional investors is to make category decisions to avoid investments outside their circle of competence.
TILT When a bad outcome causes you to be in an emotionally hot state that compromises the quality of your decision-making.
When you identify the possibility of bad luck, there are things you can do in advance to soften the impact of that bad luck. These things are called hedges. There are three key features of a hedge. A hedge reduces the impact of bad luck when it occurs. A hedge has a cost. You hope you never use it.
In a striking case of motivated reasoning, his superiors denied the evidence, insulted by the implication that their dirty hands could be responsible for patient deaths. “Doctors are gentlemen,” he was told, “and gentlemen’s hands are clean.” He lost that job, as he lost two later appointments where he introduced similar policies and got similar results. He died in a public insane asylum in 1865, at the age of forty-seven. As a final indignity, his likely cause of death was an untreated infection.
As John Stuart Mill said, “He who knows only his own side of the case, knows little of that.”
When you’re talking to somebody one-on-one, you have a simple solution to the contagion problem. Whatever the feedback is that you’re eliciting, don’t broadcast your opinion first. But this solution doesn’t scale well in a group setting. When you’re in a group discussion, you can keep your opinion from everyone, but once the first person offers their opinion, the rest of the group is infected.
Status can derive from a leadership position, experience, expertise, persuasive ability, charisma, extroversion, or even just how articulate a person is.
Group members with lower status may have different, valuable perspectives. Sometimes, they see innovative solutions others don’t see because they aren’t as anchored to the status quo.
Real accountability to a checklist means that if somebody can’t provide the information necessary for you to give high-quality feedback, you should refuse to give it—not to be mean but to be kind.
You’re going to make thousands and thousands of decisions across your lifetime, some of which will work out and some of which won’t. The goal of good decision-making can’t be that every single decision will work out well. Because of the intervention of luck and incomplete information, that’s an impossible goal. The decisions you make are like a portfolio of investments. Your goal is to make sure that the portfolio as a whole advances you toward your goals, even though any individual decision in that portfolio might win or lose.
The same is true for your decisions. Your goal is, across the portfolio of all the decisions that you make in your life, to advance toward your goals rather than retreat away from them.
If you go into your decisions thinking that you can somehow guarantee that things work out, it’s going to be very hard for you to take an open-minded walk through the universe of stuff that you don’t know. Instead, you’re going to take that walk in a permanent defensive crouch, constantly fending off the possibility that you made a bad decision or had an incorrect belief.
The word “disagree” has very negative connotations. Using “divergence” or “dispersion” of opinion instead of “disagreement” is a more neutral way of talking about places where people’s opinions differ.