In short
Generating Alpha Podcast - Episode 26: Annie Duke
Episode Overview In this episode of the Generating Alpha Podcast, host Amir sits down with Annie Duke, a former professional poker player turned bestselling author and cognitive scientist. The conversation delves into decision-making under uncertainty, risk management, and the valuable lessons that can be drawn from poker and applied to investing and everyday life.
Key Themes and Discussions
Annie Duke's Background
- Annie's journey began in academia, where she pursued a PhD in cognitive psychology at the University of Pennsylvania before transitioning to professional poker.
- She became a successful poker player, winning the 2004 World Series of Poker Tournament of Champions and earning millions.
- After retiring from poker, she focused on teaching decision-making strategies to investors, CEOs, and policymakers.
Decision-Making Under Uncertainty
- Annie emphasizes the importance of improving decision-making skills, especially in environments characterized by luck and uncertainty.
- Common mistakes made by investors include failing to recognize when to quit and not adequately assessing risk.
The Role of Quitting
- Quitting is presented as a rational and often necessary decision-making strategy.
- Annie explains that biases, such as mental accounting and sunk cost fallacy, often prevent individuals from making rational quitting decisions.
- She encourages listeners to establish "kill criteria," or pre-committed plans that guide when to exit a position.
Understanding Luck vs. Skill
- Annie discusses how to differentiate between luck and skill in decision-making.
- In high-variance environments, like poker and investing, outcomes can often obscure the quality of decisions made.
- Emphasizes the need for rigorous analysis of decision-making processes to understand one's true skill level.
Practical Decision-Making Strategies
- Expected Value: Annie explains the concept of expected value and how to apply it to various life decisions, not just in monetary terms but also in terms of happiness and personal goals.
- Mental Time Travel: Encourages individuals to think about future signals that would lead them to change their decisions, helping to create psychological distance from emotional biases.
- Social Feedback: Suggests seeking opinions from others without revealing one's own views first to gain unbiased insights.
The Impact of Over-Analysis
- Over-analyzing decisions is identified as an irrational behavior that can lead to poor outcomes.
- Annie provides a framework for understanding when to invest time in decision-making based on the potential long-term impact versus short-term emotional responses.
Daily Habits for Better Decision-Making
- Establish commitment devices to help make consistent decisions aligned with personal goals.
- Regularly practice seeking diverse opinions to counteract confirmation bias.
- Reflect on emotional states when making decisions and strive to return focus to long-term goals.
Final Advice for Young People
- Annie shares two key pieces of advice for young individuals:
- Recognize that "nothing is as important as it feels" in the moment; seek perspective.
- Make decisions that open more doors rather than closing them, embracing the flexibility to change paths when necessary.
Conclusion The episode concludes with Annie Duke's insights on the importance of rational decision-making and recognizing the interplay between skill and luck. Her journey from poker to cognitive science illustrates the practical applications of decision theory in various fields, including finance, business, and personal life.
Key Takeaways
- Understanding biases: Recognizing but not necessarily overcoming biases is crucial for better decision-making.
- Risk assessment: Establishing criteria for when to take risks or quit is vital for success.
- Continuous learning: Decision-making is a skill that can be developed over time through reflection and practice.
This episode offers valuable lessons for anyone looking to improve their decision-making skills, illustrating how concepts from poker can translate into actionable strategies in investing and life.
Written by AI. May contain mistakes. Listen to the episode to check what was said.
Transcript
Automatic transcript. May contain errors.0:01This week, I'm honored to sit down with none other than Annie Duke. decision strategist, best-selling author, and former professional poker player who has become one of the leading voices on uncertainty, cognitive bias, and high-stakes decision-making. With a unique career that spans academia, professional gambling, and behavioral science, Annie offers a rare perspective on how we make better choices in business, investing, and life. After earning her master's in cognitive psychology and completing doctoral work at the University of Pennsylvania, Annie left academia to become one of the most successful poker players of her generation, earning millions of dollars in winning the 2004 World Series of Poker Tournament of Champions.
0:39Since stepping away from the game, she's written best-selling books like Thinking in Bets and Quit, and has become a sought-after advisor to investors, executives, and policymakers navigating complex environments. In this conversation, we explore the overlap between poker and investing, the power of quitting, how to improve decision-making under pressure, and the mental frameworks that separate amateurs from experts. Annie's insights are sharp, deeply researched and refreshingly practical for anyone operating in uncertain, high-risk arenas. I enjoyed making this episode, and I hope you guys do too.
1:12Thank you. Thank you, Amy, for coming on. I really appreciate it. Well, thank you for having me, Amir. I'd like to start where I always start, at the beginning. So tell me a little bit about your childhood and upbringing and kind of how you found your way into the world of poker. Oh my gosh. How far back do you want me to go? Okay, let's see. Well, I was the child of two school teachers, although only my father actually was working. My mom was a stay-at-home mom. um and my parents had met over a game of bridge my dad was playing and somebody left and they needed a fourth and he saw this woman walking sort of up the stairs at college or I guess it was graduate school and he just said do you play bridge and she did and she sat down and that's they met um so he uh he was teaching high school she's a stay-at-home mom but he they cards still become a really big thing and the social time in our family was mainly spent playing cards although i will say not really poker and my father is not a great poker player um we were playing things like gin and hearts um a game called oh hell which was kind of mini bridge when i was a teenager i played bridge with my dad so we were we were definitely like a games playing family my brother started playing chess pretty young eventually became a master um you know we played scrabble we played backgammon that was a lot of the way we interacted as a family um so anyway I go off to college uh and then to graduate school and while I'm in college my brother drops out of college and starts playing poker in New York City so that was really my first exposure to poker um and I used to sometimes go down and kind of watch him play so I'd be able to sit behind him these were in cash games and watch him play and he would look at his cards in a way where I could see them so I could kind of follow the action.
3:23It was pretty interesting. But like at that time, there was no thought that like, oh, I'm going to go do that. But he was pretty good at poker by that time. So that started when I was in college. And then I went off to graduate school at UPenn. And I was studying cognitive science and learning under conditions of uncertainty. And about once a year, my brother would fly me out to Las Vegas because I was there on a National Science Foundation fellowship, which isn't that much money. At the time, it was$13 ,000 a year. So I couldn't really afford fancy vacations, but he would fly me out once a year when the World Series of Poker was happening.
3:59And I pretty quickly found like in a tournament, you're not allowed to sit behind somebody. So that's kind of boring. And I found that I was kind of bored. So he actually just gave me a little bit of money to go play poker on my own. And I all through graduate school, I sort of would go out and I would play once a year. But again, I'm not planning to become a poker player at this point. So at the end of graduate school. I was five years in. I'd done my dissertation work. I was out on the job market and I got really sick and needed to take some time off. And when I took that time off, I didn't have my fellowship anymore.
4:31And so my brother actually suggested, well, while you're sort of taking time off, why don't you try playing poker? And maybe you could make some money to support yourself until you go back to grad school. um and I was like okay uh and I went and started playing poker and started doing well right away and never made my way back to grad school I do want to say one thing because you're much younger than I am that it probably doesn't seem so strange for someone to become a professional poker player to you because your whole life professional poker like has been on television and so people can see it they they know that this is a thing that you do when I did that that was This was in the 90s.
5:12There was no poker on television. So generally, when people found out what I did, usually the conversation would wind its way to Gamblers Anonymous. Like that maybe I should consider that. So, you know, it wasn't like, you know, poker players weren't cool. They were like, you know, shady people who were gambling addicts as far as the as far as the public was concerned. So it was only because my brother had been playing. He had sort of gotten into poker in New York that I even knew about it. Otherwise, I wouldn't have even known it was a thing. And he's the one who encouraged me to start doing it.
5:46So that's how I made my way to that. And I think both my brother and I kind of had a knack for it because even though we didn't play poker when we were young, we played a lot of cards. And all card games have some similarities in the way that they flow. And I think that that actually really helped us develop a knack for it. You had your educational background from UPenn in kind of cognitive science. at a high level. How do you think that shaped your thinking both overall, just kind of in the world of getting a job, but also helped you become a better poker player? Well, what I think is interesting is that the first eight years that I played poker, I wouldn't say that I was explicitly thinking about the connection between cognitive science in poker um I was I was using a lot of strategies to try to de-bias myself because the problem with poker is that there's so much luck involved uh and then most of the hands that you play actually don't get to see your opponent's cards so if that's the case it's you know I don't really know like should I fold it should I not like when I won that hand did I did I actually play it well or not.
6:59It's just, it's hard to tell what's happening because of luck. Did you actually have your opponent on the right hand? These are all really difficult under these conditions. And so closing feedback loops become really hard. That becomes a really hard problem. And I did know that like, there was just a lot of bias involved in the way that people would close those feedback loops. And in particular, one of the biases that I was particularly concerned with as a poker player with something called self-serving bias, which is in any system where you can blame luck for something, you'll generally sort of lean on luck as an explanation for bad things that happen to you and lean on skill as an explanation for good things to happen to you.
7:44You can see why that would be called self-serving bias. It makes you feel better about yourself. But it's kind of wrapped up into confirmation bias that we're really kind of looking to confirm that we're pretty smart and we, you know, we're good at what we do and we want to sort of have a positive story about ourselves. And so that helps us do that. So I was aware of that tendency in poker players and really trying to work on ways to limit its influence on me. But I can't say that I was doing that explicitly. It's kind of interesting. It's like I left graduate school and then I was playing poker.
8:19And I wasn't really thinking about like the connection between the science that I've been studying and this game that I was playing. What happened was that in 2002, I got asked by a hedge fund to come talk to their traders, their options traders about risk. And they wanted me to do that through the lens of poker. And what I actually talked about to them about was not risk per se, but risk attitudes. How do you approach your decisions about how sort of risk-seeking you're going to be versus risk-averse you're going to be dependent on the path that you've been on? So I was just pointing out, and this is a well-known finding in science, that if you've been losing a lot, you'll tend to get pretty risk-seeking within a game.
9:12And then if you've been winning a lot, you'll tend to become pretty risk-averse. meaning I'm going to gamble a lot if I'm losing because I'm trying to get my money back and I'm going to prefer not to gamble to protect the win that I already have and this is a central part of something called prospect theory which is from Daniel Kahneman and Amos Tversky in fact Daniel Kahneman won a Nobel Prize for his work in that so this is a central finding that has been known in science since 1979, very well replicated that this is kind of what people do. So that's actually what I talked about. And I would say that that's the moment where I was like, oh, you know, these two things are actually really related to each other.
9:57And as you think about the scientific findings, wow, you're like really seeing all of these at the poker table. And then you see that, first of all, it's a good sort of reinforcement that these things are really occurring in real life and they are really a problem. But then also. The good poker players have come up with interesting strategies to try to mitigate the effects of these things that I think can actually inform some of the science. And so the two disciplines can kind of like really reinforce each other and inform each other. And they're sort of synergistic, right? So that was really the first moment that where I went, oh, in this very explicit way.
10:43I mean, I did realize I'd been sort of thinking about it all along because I was trying to sort of de-bias my, the way that I was sort of trying to figure out what was going to make me a better player and why certain things happened at the table and that kind of thing. But it wasn't until I got asked to speak about it explicitly that I went, oh, right. And then I spent the next decade developing a suite of talks that were a variety of different ways in which poker and cognitive science could inform each other. And that speaking eventually became my first general audience book, which was Thinking in Bats.
11:19And then obviously I've written some since then, but that was sort of the genesis of that. circling back to kind of your time in poker obviously you observed many types of players and poker is a game where there's risk involved uncertainty luck all kinds of things are there any kind of like types of poker players that you think reveal things about types of humans like people that are more risk seeking risk averse people get emotional at a certain time period yeah so I think that there's a couple of things the first is that no matter who you are there are commonalities at the table that just reveal some of the things that we do wrong.
11:55So this idea that if I'm in a game of poker and I'm losing, two things are going to happen. One is that I'm going to become pretty risk-seeking. I'm going to gamble more. I'm going to play more hands. I'm going to want to raise more. And I'm going to be much less willing to quit the game. So the way that I think about it is if you do the thought experiment of you're losing and you were supposed to go to dinner with someone, how likely is it that you make an excuse that you can't go, right? And then think about the opposite. Like you're winning a whole bunch of money and you have that same engagement.
12:35Like how fast are you getting your chips off the table? They're like, oh, I'm sorry, I have to go to dinner, right? So when you're winning, you'll tend to, you'll have this tendency to want to bring your risk, to bring the volatility down. So you're, you often get picker about what hands you play. You're not, you aren't necessarily as aggressive and you're faster to get out of the game. So it's like the opposite of the sunk cost effect, right? The sunk cost effect is if you're losing, you try to get your money back and it stops you from stopping. But when you're winning, you actually want to sort of close that ledger out, right?
13:16You want to just lock the gains up. And so you want to kind of bring your risk to zero. So there's this generality about human beings that's revealed. It doesn't matter who you are, and it doesn't matter what your general sort of baseline risk attitude is. You're going to see these fluctuations from what, if I take what your risk attitude is under just like starting circumstances, when you're losing, you're going to be more risk-seeking. when you're winning you're going to be less receiving another thing that poker really reveals it's like a generality that's general to human beings is that if you've been on a winning streak so this is different right I was talking before like you're in an individual game and you bought in for like two hundred dollars and now you're winning like three hundred and fifty dollars and you kind of want to leave the game because you want the money but let's imagine that you've been playing, you know, you've just been on a crazy winning streak for three weeks.
14:10And every time you play, you just can't, you just win. You'll actually, the next time you play, be too risk-seeking because you'll, you're sort of like where you think you sit in terms of like how much you're at, like how good you actually are is going to be wrong. It distorts you. And likewise, if you've been on a long losing streak, then if you come in the next day, where you start is going to be risk averse. So that it flips, like once you sort of get out of the game and it's been going on for a while and blah, blah, blah. So that's a generality. And then self-serving bias is a generality. So you, poker reveals a lot of things that are very general about human beings.
14:51And then what you do find is with specific human beings, and it becomes part of your model of the player that you're playing against. Some people are naturally more risk seeking. some people are naturally more risk averse so the risk aversion can uh kind of it generally will manifest in on this dimension which is called tight or loose so um that's one dimension that you can see it on so tight means that you don't play very many hands and loose means you play a lot of hands. So obviously the more hands you play, the more risk you're taking on, right? And some players are tighter and some players are looser.
15:40So that's one dimension that you'll see that people really vary on. And then the other dimension is aggressive versus passive. So the question is, if you've chosen to play a hand, do you play it aggressively, meaning you do a lot of raising, or do you play it relatively passively, which means that you do a lot of calling. And the more raising you do, the more risk you're taking on. So people really, really vary on those two dimensions. And what I think is really interesting is that I can take a pretty well-educated guess of where you're going to sit on that scale, depending on some interesting things like what time of day and what day of the week it is, how you're dressed, what your age is, just a whole bunch of stuff like that where I can never have met you.
16:41And where I'm going to start person A on that passive, aggressive, loose tight scale is going to be different than person B. So you can think about like, let's imagine that it's 11am on a Wednesday, and like a locals casino. And the person that I'm playing is retirement age. I'm going to put them in the much tighter range, and probably more passive range, right? So that's where I'm going to start them. If it's two in the morning, and it's someone who's there, you know, in Vegas, and maybe they've had a few, and they're on vacation, I'm going to put them on the more aggressive range, and the more the looser range.
17:24The important thing about that, though, is that as I start to see them play, I need to start adjusting immediately, right? So it's I'm just starting them in different places, but I'm not sticking there. It's the difference between like an archetype and a stereotype. So I'm not stereotyping them. I use different archetypes to have me start in a different place. And then I update from there. But so I think that's something that's kind of interesting is that you can actually tell a lot, like people say you can't judge a book by its cover, but you can a little bit, right? Like at the poker table, you certainly can.
17:56And, you know, and then you, but the key is that, so it doesn't turn into a stereotype is that you do actually start to read the book, right? I kind of want to focus on, I want to now shift to your work over the past 10 or 15 years, because my podcast is focused on kind of finance. And I think there's a lot of lessons people can draw from your work and specifically decision-making that can aid them both in investing and overall in life. You've done a lot of work on kind of the difference between luck and skill. How can one distinguish between luck and skill, just both in investing, but also in life?
18:33that's a mighty broad question okay so let me um
18:41okay so let's just start with this there's some things where you can just measure the volatility right um yeah i track the vix you'll know how much volatility there is in the market right like that's gonna be a pretty good indicator of how much volatility there is but some things are going to be more volatile than others. Now, it's going to vary around whatever that base rate is, right? Like poker has a certain amount of vol associated with it. But if I'm playing in like a looser, more aggressive game, it's going to be higher vol. If I'm playing in a tighter, less aggressive game, it's going to be lower vol.
19:19So generally, you should know something about sort of how much luck am I expecting to see in So like an example would be backgammon, the luck is kind of known, right? Like because it's dice, right? So we know how much volatility there's going to be in a game of backgammon. So that's one thing, right? But the more important question, I think, is in a world where outcomes and decisions in the short run are not perfectly correlated, in other words, outcome quality and decision quality, how is it that you disentangle what is because of luck and what is because of skill? And that's actually hard to do.
20:01So let me explain what I mean. So if you're playing chess,
20:10you don't need to see the game that's played to know that the winner made better decisions than the loser on average. And that's even true if like the loser mostly made better decisions, that but then made a fatal error you can safely say that well that means they made worse decisions because they made a fatal error right um and the reason for that is that chess doesn't have dice right like where the the pieces move because i move them like that and then they stay there until i move them again and you're not like randomly getting extra pieces like you know you don't have exploding bishops or something like that, that are stochastically exploding or something like that.
20:51So, so that's, and, and all the information is known, like you can see where all the pieces are. So I don't need to, I don't need to actually have seen anything in order to know that the winner was the better decision maker, but that's not true of most things in life. So I think it's kind of hilarious that people are like, they're playing 4d chess and it's like, okay but like chess isn't actually as hard as most of the other things that you do um so how do you disentangle it one is you can get a very large sample size so that's one way that you can do it right if you want to figure out if a coin is fair and you you can't weigh it right you don't have access to be able to weigh it you're just going to have to flip it enough times to know that the coin is actually fair that it's going to land 50 50, right?
21:39And that's going to be some number of times that you're going to have to do that, right? So you could gather a big sample. The thing about that though, is, and one problem that we have as decision makers, by the way, is we tend to draw conclusions before we have a large enough sample. So when I talked about poker players who like are on a winning streak, and then, you know, after they've been on a winning streak for three weeks or something like that, they start changing the way they play because they're overestimating what their expected value is. That's because they're making judgments off a sample size that's too small for the volatility.
22:15Right. It's not the sample size isn't overcoming the volatility to give you the truth. And that's a problem that we have as decision makers is we tend to make decisions. We tend to make judgments about how good or bad we are at something or what the meaning of something is on very small sample sizes, sometimes on just anecdotes, right? This is just a problem with the human brain. So, so, you know, we want to be very careful about like how many instances are actually looking at given the volatility in the system. So that's one thing you can do, but you want to actually know before that. So remember when I said the thing about chess is that you can tell that without seeing the moves.
22:52What's better is if you can see the moves. So if I can see the moves, then I can actually tell something about decision quality. So the way that you actually can figure out what's going on is to make decisions in a way that is creating an artifact of the decision of some sort so that you can understand what are the things that we knew at the time? What judgments did we make about those things? What forecast did we make about what we thought was going to happen in the future? And then across some set of decisions, you can start to close that feedback loop to figure out, well, how good actually am I at these decisions?
23:34And you have to be pretty rigorous about it. Otherwise, it's actually really, really hard to figure out, like, am I actually good at this? How much skill do I have? How do I actually improve? so it really comes down to like how rigorous like how systematic are you in the decision making process itself you spoke a little bit about expected value can you talk a little bit expected value and how to kind of gauge expected value of your decision yeah so expected value is it's hard for people because it's an abstraction. So let me just say any decision that you've ever made in your life is probabilistic.
24:19What that means is that there's pretty much no decision that you make, which is going to, you can guarantee what the outcome is going to be. Most of the decisions you're making will have some set of outcomes that are associated with them. So you're studying for a test. You could get an A, a B, a C. Each of those is going to have some probability, and then your actions can affect the probability of those outcomes occurring, but you can't guarantee an A+. Right? All you can do is increase the probability that you're going to observe the A plus. So that's what I mean by probabilistic in nature. It's like, we know this, like, if you're choosing a route to go somewhere to work or whatever, you're making some forecast about what the most likely time to travel is.
25:16But like, there could be an accident that makes it so that, you know, you're very late, right? Like that's what I mean by like probabilistic. It means it's not a sure thing. it's going to occur some percentage of the time. So when we're making decisions, what we're trying to do is make the decision that on average is going to advance us toward our goal the most, given the resources that we have and the risks that we're willing to tolerate. So what we're trying to do is gain ground toward our goals. And the way that we judge how much ground we're going to gain toward our goals on average is by calculating an expected value, which is the probability of each of the outcomes that we would expect to observe given a decision that we might be making, an option that we're choosing, multiplied by the amount of ground it will cause us to gain toward a goal or the amount of ground it might cause us to lose toward a goal.
26:16So let me give you a very simple example. So we can simplify it by saying, what if our goal is to make money. Very easy way to think about it. And I'm trying to decide, you know, Amir has offered me a bet. We're going to flip a coin. Let's imagine that I've weighed the coin. So I know the coin is fair. So I have no question about that. And he's saying to me, okay, let's make a bet. And I'm going to give you$2 if you get heads and you have to pay me a dollar if it lands tails. And my question is, well, is that a good bet for me, right? My goal is to make money, not lose money. Is that going to cause me to gain ground toward this goal of making money?
27:04And so I can do exactly what I said. I know what the payoffs are, right? Like I'm going to win$2. I'm going to get$2 closer to my, whatever my goal of making money is. If it lands heads and I'm going to lose a dollar, I'm going to lose ground toward my goal. if it lands tails. And I know that each of those outcomes is gonna occur 50 % of the time. So expected value is just a weighted average. So I would take 50 % 0.5 times$2, which equals plus a dollar, right? So I'm gonna win$2 half the time. So that's a dollar. And then I can take a loss of a dollar times 50%, because that's gonna happen half the time, which is 50 cents.
27:47And then I take the dollar that I'm winning and subtract the 50 cents that I'm losing. And the expected value for that bet is 50 cents, meaning for every dollar that I have risk at risk, I'm going to make 50 cents. How can you apply that to just like not not necessarily in investing, but just general life where not everything is like numerical and not everything's statistical? So first of all, let me just say something about this this sort of coin flipping problem. yeah one of the problems that people have with sort of thinking and expected value is that there's no coin flip where you give me 50 cents right so i'm making 50 cents on each coin flip but you never give me 50 cents yeah you either give me two dollars or i give you a dollar um and this is actually relatively hard for people to wrap their heads around not surprisingly right because it is an abstraction right this is what i'm sort of my i'm expected to make over the long run so what i know is that if we did 10 ,000 coin flips, I should walk away with$5 ,000 over the long run.
28:50But we could make the bet three times and I could be down$3. That wouldn't be that surprising, right? That would be 12.5 % of the time, I'll lose$3, right? We agreed to do it three times and I walk away losing$3, even though I was making 50 cents on each flip. 12.5 % of the time, I'll walk away with three dollars even though i was expected to only walk away with a dollar 50 right so like both of those things can happen so this becomes like it's interesting because it's hard for people to think about even if we're talking about something as simple as money but then they're like wait a minute but what how would i do this in another way um that doesn't have to do with money like how would that apply well we can think about a simple example like the traffic right right what's my expected drive time.
Read the full transcript
29:43That's an expected value. And it's like, I could look at what the distribution of outcomes are. How long does it take me over 40 minutes to get there? And I can look at historical numbers, right? Say, how long is that? And that gets me to the expected value. But again, like if the expected drive time is 31 minutes, how many times is it exactly 31 minutes, right? Just that's my expected time over the long run, right? If I took an average of all my trips, it would average out to 31 minutes. So that's another way we could think about it in terms of time, but you can think about it in terms of things like happiness.
30:18And this is a really important idea around really great decision-making to say there are things that are qualitative in nature. What happiness means for you is different than what happiness means for me. But we can both judge like on a scale of zero to a hundred, how much, how happy do you think this would make you. And we can both make a judgment about that. Now, our answers might be different. That doesn't mean that one of us is wrong. One of us is right. It means that we value different things, but we can make that judgment. Right. So we could think about like on a scale of zero to 10, if you got to go to Rome for a week, how happy would that make you?
30:58If you went to Thailand for a week, how happy would that make you? If you went to the Caribbean for a week, if you went to the ruins in, you know, Mexico, or South America, like how happy would that make you? And we may have different answers, because you may be a beach person, and I may hate the beach. That's okay. So but you can rate it. And then you can, then you can start to calculate an expected value. One of the things that you can do is like, you could turn it into a currency and there's a few different ways to do that the rating kind of turns it into a currency but you could think about like it's really something called expected utility how much use how much value are you going to get out of the thing um and you can create a currency it's called utles people go go look it up um but another thing you could do to get you there is how much would i pay for someone not to take this away from me so that's like an interesting way to get to it right if I had a trip to Rome, how much would someone have to pay me for me to give it up?
32:05Yeah. And that's another way to get at that sort of quantification of things that feel mushier. But I think the thing that people need to understand is that for you to figure out where to go on vacation, somewhere under the hood, this is happening because you have to balance out. Like, do I want to go to Machu Picchu or do I want to go lay on a beach in the Caribbean? two really different vacations right so if you're comparing those you are actually already exploring your values uh you are in some way like quantifying this idea of like how happy do i think that each thing is making me so this just asks you to make that explicit and say how happy do i think it's going to make me and with what probability will that occur if i think about all the different vacations that i could take and that helps you get to an expected value yeah i think that that kind of idea of how much would you pay me to not do this is interesting, bringing it back to that kind of quantitative point.
33:02Some of your work, which I'm particularly interested on is specifically in when you're trading or when you're investing, knowing when to quit is extremely powerful. So what role does, have you seen like ego and identity play in knowing when to quit? And how can someone kind of limit those and be able to have a baseline, just like a kind of rational metric of knowing when to quit and decision making? Yeah. So here's the thing about all of the biases and the failure to stop is a really big one it's why I wrote a whole book about it um we particularly anybody over the age of 25 and many people under the age of 25 stick to things too long um and that's true in trading as well and it's particularly true if you're in the losses going back to the poker table right it's like when you're losing you don't want to quit because there's this concept that we can think of called mental accounting.
33:59And so let's imagine that your benchmark is even, right? Like you bought the stock at 50. So that's now your benchmark. And it's trading at 40. Your mental account is in the losses. And when we're in the losses, we don't like to close the account. We don't want to close the account in the losses. It doesn't matter if you own 20 other stocks, which of course would be the rational way to think about it. It's portfolio theory, right? but your only thing, you're only like, I don't want to sell this stock is like, then I won't get my money back, which is stupid. The reason why we know that that's a mental accounting problem is that if you buy the stock at 50 and it's now trading at 75 and then it goes down to 65, you will also feel like you're in the losses, even though you're not, you're up$15, right?
34:43Okay, so we can think about those types of biases, right? Like sunk cost and this mental accounting problem and endowment. And then those intersect in a really bad way with a lot of these things that have to do with identity, right? So we like to be seen as consistent over time. We want to feel internal validity means we want to feel like we're smart and we're good actors. And so like, if we lose or we stop something that we started, then like, what does that mean? Like, did we make a mistake when we started it? So that doesn't feel very good to us. And then we also think other people are going to judge us harshly for those decisions, even though in general, they don't.
35:20um so we have this sort of intersection of like the way that our brains just kind of like process this stuff like being in the losses and then how that intersects with identity how do we view ourselves how do we think that other people are are going to view us and that uh it all creates sort of a failure to stop on time and a failure to stop on time is really bad because if i'm holding a stock that has a lower expected value than other stocks that I could be in. Or maybe it has a negative expected value, right? I'm getting hit two times. One is I'm not earning as much as I could, or I might be losing money on the path that I'm on.
35:58But the second thing is I can't move those resources over into something that would be better. And that's true, whether it's like money that I've invested in stock, a stock that I ought to get out of so that I can move it into a better stock. But it's also true, like if you're in a job that you hate, not only do you hate that job, but you can't go get another job that you're really happy about. Or if you're in a relationship that isn't going well, it doesn't matter what you're doing. It's like there's always this problem of the cost of being on the path, but then also the opportunity costs that you're incurring.
36:29So we want to get ourselves into a place where we're actually willing to stop things sooner, right? That's generally what the goal is because most things are stacked against you stopping. The issue is that knowing about it doesn't help, but it really feels like it will. So I've had a lot of traders say things to me like, oh, yeah, yeah, you know, I know about all that stuff, right? Because they do. Like they've read Thinking Fast and Slow by Daniel Kahneman or something. They're like, I know about all that stuff. I know about those problems. So what I do is every morning I wake up and I think, if I weren't holding these positions, would I buy them today?
37:10Okay, well, I agree that that's the question that you should, like, it's true, right, that if you wouldn't buy it today, you ought not be holding it. That is true. And that's actually the test of whether you would be the test of whether you are quitting. Well, if I were an omniscient, totally objective being, right, and I were staying in a position that I would not buy today, less transaction costs, then I am subject to some sort of bias that is causing me not to quit because those two decisions should align with each other. If I wouldn't buy it today, I ought to sell it. So they've got the sort of like in that sense, the right question.
37:54The problem is that that's only the right question if your answer is going to be a rational one. And once you're actually already in the position, the answer isn't going to be a rational one. So then what happens is that you have an irrational answer to that question, but you think that you were rational about it. And then your confidence that you're actually correctly in the positions that you're in has gone up, even though the accuracy of that decision has not. Okay, so that's bad. And by the way, this is true of almost any bias, right? You can't like think your way out of confirmation bias.
38:30You can't think your way out of, you know, availability bias. Yeah. You know, you name a bias. You can't think your way out of self-serving bias. Like, it's really hard to do. So then the question is, okay, so what do you do? And the answer is that you have to do some relatively sort of structural and brute force things in order to get yourself to do it. And most of those things have to do with saying, you know what? in the moment that I have to face down the decision about whether to quit, there's too much going on for me. There's not just the problem of like, you know, I might be in the losses in my mental accounting, but also like my identity is wrapped up in it.
39:13I don't want to feel like, you know, a dummy that I got into the position in the first place. I want to get my money back, like all those things. And it's, and you just have to realize like when you're facing that decision down, you're just not going to be good at it. So what you're trying to do is create distance from the decision combined with some pre-commitments in order to get yourself in a spot where you'll actually stop more rationally. So there's two strategies. One has to do with something called mental time travel, which gives you psychological distance, which is to say, like, let's imagine that you're putting a position on, you could say, what are the signals that I could see in the future that would cause me to take this position off?
39:58The simplest version of that would be a stop loss. Simple, right? I buy it at 50. If it's training at 40, I have to sell it. Okay, that's really, really, really simple. Very, very brute force. Probably better to link it back to your thesis. But if that's all you did, if all you did was put in a stop loss and actually follow it, which is important, that would be better than nothing. But what's better is to say, look, I'm getting into this position because I believe something is going to be true about say interest rates in the future. And if interest rates are above a certain benchmark, which I'm predicting they're going to be a high enough percentage of the time to make this a good bet, then you know this is a position that I ought to like being in but if they go below that then this is a position I ought not like to be in and what happens to us is even though that's our thesis in the beginning right as we hold that position if we get into a losing position even though interest rates aren't where we want thought they were going to be we'll rationalize reasons to stay in the position.
41:10But if you actually make it explicit and you write this down and you say, okay, I'm going to put this position on by X date, right? Or any time that interest rates fall below this certain benchmark after this certain date, I must sell it. If interest rates are above, then you could say if they're within a certain band above that benchmark, then I ought to hold it. And if they get above that, maybe you would even put more risk on, right? So you're actually committing, you're saying, what are the signals in the future that I could see that would tell me that this is a trade that I ought to take off?
41:50I should take risk off. This is a trade that I ought to hold. This is a trade maybe where I should put more risk on. So you're actually sort of thinking about that now, but thinking about something that's going to occur in the future and then committing to those, what I call kill criteria, that you're actually going to act in a certain way. That really increases the chances that you're going to quit. And then if you actually tell somebody else about it who will hold you accountable to it, then that really, really, really increases the chances that you'll quit. What I think is interesting is that when I tell people about this, they're sort of like, but what's the difference?
42:27Either way, like, let's imagine the interest rates, you know, go up 50 basis points. They're like, in both cases, interest rates went up 50 basis points. Like, why does it matter? And it just turns out that it really matters if you decided in advance than if you're deciding on the fly when that thing occurs. It just really makes a huge difference. So it's such a simple thing that you can do, but it will actually get you to de-risk when you ought to more often. We've been talking about kind of rationality and making good decisions. Is there any downside to kind of being too rational and overanalyzing things well overanalyzing things isn't rational yeah so no um so actually so one of the things that i teach in my classes is when should you actually take your time with a decision and when should you not uh and believe it or not like so the way that my classes went so so i teach a class i i taught at wharton for many years and then um And then now I teach on a platform called Maven, which is like a direct to student.
43:39And it's a six session class. And I spend the first five sessions teaching them like all this structure and all this process around decisions. And then the last session is, but mostly you should go fast. So people actually do not make rational choices. Like overanalyzing is by definition an irrational thing to do. It means that you're putting too much effort into a decision that ought not, you know, deserve it. So there's two really big ideas for sort of how you figure out how much effort. The first is long-term impact, not short-term impact, but long-term impact. So you can think about like, do you know anybody like who they're in a restaurant and they're just taking, they take like forever to try to decide what to work.
44:24Right. Like forever. Right. And the thing is that if you said to them, imagine it's a year from now, like imagine that you ordered whatever you ordered this chicken and it was pretty crappy right like you didn't even finish it because it was like overcooked and yucky so and if I catch you in a year and I ask you like hey how happy have you been over the last year how much effect does the chicken have on it like it's zero right in fact if I catch you in a month that the bad chicken still has zero effect on your happiness and actually a week later it doesn't it might have some effect that day but so not a week later, not a month later, not a year later.
45:02It's a very low impact decision. Yet people overanalyze, right? Like they freak out. And the reason is that they're going to get an answer really fast and they're afraid that they're going to make a mistake, right? As if ordering off the menu is like two plus two equals four and they're afraid that they're going to answer seven. And then they're going to find out that they were wrong. That doesn't make any sense. It's like you haven't had the chicken yet. You're making a guess based on your preferences and the description on the menu. But the more important point is this, it doesn't matter. It doesn't matter to your long term happiness.
45:35So like, divide the menu into stuff you like stuff you don't like. And then I don't care how you decide, like ask the wait staff to bring you something or flip a coin or ask a friend to choose. It doesn't matter. Just choose. Right. So that's the first thing is like Like this long, this idea of long-term impact. So you can think like, don't take that much time deciding whether to go on a date with someone, but take a lot of time deciding whether to marry them. That would be a good thing to do, right? Don't take a ton of time trying to decide what intern to hire, but take a lot of time trying to decide what CEO to hire, right?
46:14That's an impact question. The second thing actually has to do with what we were just talking about, which is quitting. um the lower the cost to quit the less effort you ought to put into a decision and the reason is that one of the best ways to mitigate observing sort of bad outcomes like having things kind of not go your way is to stop what you're doing so uh the lower the cost to get out of it the faster you can go because you it's easier to mitigate the downside um once you sort of observe that now of course that means that you have to be good at quitting so you have to do kill criteria in this mental time travel and like get other people to help you.
46:52And that all helps you to distance from the decision. But, but that's the other thing that matters. So if we go back to the dating versus marrying, it's really easy to just not go on a second date with someone like that's a very easy thing to quit. But marriages are very hard to quit. They're very expensive, both like money wise, emotionally, you know, disruption and all that stuff. So take a lot more time deciding who to marry and very little time deciding who to go on a date with because the date is more quittable at a much lower cost. You can leave in the middle of it. You can be like, yeah, I got somewhere to go, right?
47:25So the sort of over analysis, people think about it as rationality, but it's not rationality in the sense that a cognitive science would think about rational, right? Does this decision actually deserve the time that you're spending on it? so what you're talking about is like what about people who are like over analytical i think but that's not right like as soon as the word is in there it's not rational yeah i have i have one more question before we kind of move on to some of your advice um but we've talked about all this are there any daily habits or practices that people can put into play that kind of make you a better decision maker over time, even if it's just like a small daily thing you do every day?
48:16Yeah. Well, one is prodigious use of pre-commitment contracts. So really kind of think about like, it's really good to think about what are your goals and then think about like, what are the barriers gonna be? Or like, if I imagine that I don't achieve this goal, like what do I think happened? And then actually commit to every day making sure like that you're doing the things that would avoid, uh, that would avoid not actually, you know, achieving your goal. Right. So, um, you know, one of the things I love is this idea of a category decision, right? So, um, let's imagine that you're trying to sort of allocate your time toward fun and studying, make a category decision, which is like, I don't play video games from 6 p.m.
49:16to 9 p.m. on school nights. Why does that matter? Because then you've actually in advance made a decision that that's not the type of person that I am, in the same way that someone might say, I'm a vegan, right? So you're sort of taking video games as an option during that time off the table. So I would say it's really important to like be very prodigious in your use of those types of commitment devices. Making those types of advanced decisions are really important. The second thing that I would say is, and this is something you can do every single day. If you really want to become a better decision maker, which partly means that you have to get better information from people, other people are very helpful in making good decisions and helping you make good decisions as long as you're actually hearing what they think so you can think about it this way like I have all sorts of biases I'm going to tend to want to think about the world in a way that like confirms what I already believe so are you but the things that you believe are the different different than the things that I believe.
50:28So what happens though, is that when we're talking to other people, we tend to tell our story and then ask them what they think. And that's actually not helpful for correcting where we might have things wrong. So one of the simplest things you can do is ask people what they think without telling them what you think first. And if you make that a daily habit that you put into practice, you'll be surprised how much more like really interesting opinions and information you get from other people. I remember having a conversation with my son when he was in high school and he was telling me he was in a fight with a friend of his and he was telling me how all his friends agreed with him that this person was a jerk.
51:14And I said to him, well, let me ask you a question. did you say to those people what do you think of so-and-so which would be neutral or did you say man he's such a jerk he's you know did this and this and this and what do you think don't you agree because in the second case of course they're saying they like you didn't actually find out what they think and of course he was like okay fine and I was like well what do you think would have happened if you had just said what do you think of so-and-so you know and he kind of realized oh oh, well, then I might actually get different, like I would have gotten very different answers to that question.
51:49So I think that that's actually just like a really important, I think it's a really important skill to start to implement in your daily life. And then the last thing that I would say is that I think that it's really easy to get very upset about things. And, you know, and I understand why, right? It's really upsetting if you get a bad grade. It's really upsetting if you lose money in the stock market. It's really upsetting if your boss is a jerk to you that day. It's really upsetting. But first of all, we're more likely to get more bad outcomes in the future if we start making decisions when we're in a hot state, when we're emotional.
52:30It's actually part of what the pre-commitment contracts help you with is they take you out of those hot states when you're making decisions. But one thing that I would say is that when you're in those spots, you should actually really think about, well, what is my goal here? And how would I actually accomplish that? Because it's important to get past the emotion and sort of wanting to satisfy and relieve that emotional state in the moment and get back to how am I gonna gain ground toward my goal? So I'll give you another example. I'll just pick on my son. He was really mad because he got a bad grade on his test, a test that he took and he did the self-serving bias thing right the test was really hard okay that's not his fault right that's like a poker play saying I lost because of the deal um the test was really hard the teacher doesn't like me you know I'm sure you've heard all this before he there was stuff on the test that he didn't teach in class you know so on so forth so it's all that stuff right none of it was like I didn't study or whatever so he was really upset and he was really like complaining to me about it.
53:34And I was like, well, do you, what's your goal here? Like, do you want to try to do better in this class? Do you want to maybe try to get a chance to retake the test or make sure that on the next test you're doing better? Or do you want to just sit here and like, tell me everything that went wrong. Right. And I did this in a very nice way. Um, and you know, he was like, no, I want to do well in the class. And that's why I think this was so unfair and I was like okay well uh what are the things that you could do that would help you to actually achieve that goal and we talked about it you know and I was like well did you go see the professor before the class before the test no what do you think would happen if you did that in the future oh well that might help right um what if you went to him what do you think would happen if you went to him with this test and you actually asked him to go through it with you to figure out what you did wrong, right?
54:26So all this is doing is, and this is a really important practice is like, okay, you're going to feel your feels, right? But then get refocused on your goals and that's going to help you to get into a calmer state that's going to allow you to start thinking about what are the things that I would actually need to do in order to cause me to gain ground toward the thing that I actually want instead of just getting mad and being reactive. Yeah, I think those are all extremely powerful and I think I'm going to start implementing them. But final question, I ask every single one of my guests at the end of each episode, if you had to give one piece of advice to a 15-year-old high schooler, college student, whatever you want to think of, a young person today, what would it be?
55:07Well, I'm going to steal from Daniel Kahneman and it relates to what I just talked about. Nothing is as important as it feels when you're thinking about it. I think that it's so important to realize that sort of two things that are related to that. You know, one thing is that you, in the moment, it just feels so big. But if you can get past that moment somewhere in the future, you'll realize it's just a small thing in, in a series of stuff that's going to happen to you. It's kind of like the chicken that doesn't matter at the restaurant with my kids, when stuff would happen and they would just feel like it was so awful.
56:00And the worst thing I would say, tell me the story you're going to tell when you're 40 and you're at the Thanksgiving table, you know, and it was always a funny story. This thing that felt so horrible and so awful. And like everything was ending when they would tell the story of like, how are they relaying that, you know, 30 years from now at Thanksgiving, it would actually turn into a funny story. So I think that's something that's really important to remember is like, it's hard to get things sort of into context of the scope of time. And I think it's particularly hard for younger people because they haven't had as much experience sort of being able to look back at other things that haven't mattered so much.
56:41But the more that you can do that, I think the happier you're going to be. because it's like that the stuff that feels so big in the moment isn't going to derail you in the same way. And I think that you're going to be happier. I think the other thing that's important to kind of understand is that the world changes. And so, you know, you should try to accept that. And what you should be doing when you're making choices is make choices that open up the most doors for you and not choices that close a lot of doors. So this is like, increase your options. Don't decrease them. So like one of the things that I love, it's a funny statistic.
57:26I don't know about now, like in the age of AI and all that, but let's, it was, you know, 10 years ago, the college major that kept the most opportunities available to you was philosophy and the PhD that created the fewest opportunities available to you was philosophy that I saw was kind of funny but you know if you're thinking about like as you go to college like what college has the the sort of broadest range of things that I could study of those things that I could study what thing is going to open the most doors for me right and you should always prefer the thing that's going to be a door opener, right?
58:09As opposed to a door closer. And then understand that it's okay to switch, right? Like when you go to college, you haven't been at the college before. If you don't like it, you don't have to stay there. You can transfer. It's not a big deal, right? If you decide on a major and you don't like it, switch. It's okay. It's actually a better thing to do. Like you're young, you have the time to do that. So I think like, you know, start being good at mental time travel and then start realizing like it's okay to just switch and do different things. And that's actually going to make you a more sort of like interesting and well-rounded and hireable individual.
58:50So always be thinking about opening doors and always be okay like saying I'm going to walk through that door instead of being in the room that I'm in now. And I'm going to go see what's in that other room. And I just think you're going to be a happier person. I think that's great advice. Thank you for coming on. I really appreciate it. This was a lot of fun. I hope you enjoyed it. I did. Thank you so much for having me. I love what you're doing. Thank you. Appreciate it.
From the publisher
This week on Generating Alpha, I’m joined by Annie Duke — a former professional poker player turned cognitive scientist, bestselling author, and one of the world’s foremost experts on decision-making under uncertainty. Over the course of her career, Annie has helped everyone from investors to CEOs to policymakers improve how they think about risk, probabilities, and knowing when to double down — and when to walk away.
Annie’s journey is anything but traditional. She left a PhD program in cognitive psychology at the University of Pennsylvania and went on to become one of the most successful poker players in the world, winning millions — including a World Series of Poker championship. After retiring from poker, she dedicated herself to teaching others how to make better decisions in environments where luck and skill collide.
She’s the author of multiple bestselling books, including Thinking in Bets, How to Decide, and Quit, which have become required reading for decision-makers in finance, business, and beyond. Her work sits at the intersection of psychology, game theory, and behavioral finance.
In this conversation, we explore how to make smarter decisions under uncertainty, the common mistakes even the smartest investors make, how to know when quitting is actually the most rational choice, and the frameworks she’s developed that anyone can apply — whether at the poker table, in markets, or in life.
