Transcription
What makes a great bet?
Every single decision that you make is actually a bet because every decision is made under uncertainty. A good bet is going to be one that carries positive expectancy. Was it a good decision for LeBron James to be tall? No, that was just luck. Like every single person who has ever made a trade at some point has made a terrible trade and they've won anyway. Poker, which is just trading. I get my money in the pot and I have an 8% chance of winning. Well, guess what? That means I'm going to win 8% all the time. And the thing that I don't want to do is win that hand and then say, "Well, that actually that was a really good decision." Because it wasn't. It was a horrible decision. If you go through a green light and someone comes through the other direction and t-bones you, it doesn't mean it was a bad decision to go through the green light. If I can go through a green light and get in an accident and go through a red light and not get in an accident, I hope I don't work backwards from those outcomes and decide that I ought to start going through red lights and stopping at green lights. What we can do is get it better. Right now, you're grinding a pretty big edge over the other version of you. What a good bet is not is one that wins.
Annie, thanks so much for coming on the pod.
Thank you for having me. I'm excited to be here.
What makes a great bet?
Ah, okay. So, well, first of all, let's define what a bet is. How about that? So, a bet is um when we invest resources, which doesn't have to be money, time, attention, effort, could be money. uh into an option. So we have different options that we could in uh invest those resources into um uh with uh some sort of forecast involved of what the expected value of the outcome is going to be. Um and what we're trying to do is invest in options that have the highest expectancy given whatever our risk tolerance is. Um, honestly, interestingly enough, risk is a relatively small part of that equation, assuming you're you're within a margin. Um, so, uh, we could think about that as investing in a financial instrument where it's pretty clear you're making a bet, right? So, it's it's uncertain what the outcome is going to be. It's a forecast. There's some range of outcomes that could occur. Each of those outcomes have some probability and payoff associated with them. we can we can take a weighted average of what those things are to calculate the expected value. Um but we're guessing about that because uh we're only considering the options we're considering. There may be other ones we don't know about or we haven't thought about. Uh we have limited information because we're not omnisient and we don't have a time machine. So we can't actually know in advance which which of those possible outcomes we're going to observe which which makes it a bet.
So, with a financial instrument, I have a variety of stocks that I could invest in. I have different trades that I could make. Um, and I'm trying to pick the ones that have the highest expectancy. Uh, and obviously here we're measuring that by money. Um, but what job you choose is also a bet, right? Like you have different things that you could apply to that takes time. Uh, different jobs that you could take. Um and the way that we calculate expected value there is when we take a job we have different goals. Um so part of that goal would be to make money. Um part of it might be fulfillment uh connection with your co-workers, autonomy, creativity, whatever it is. And you're different people will have different waitings on those things. And then they're trying to choose the job uh that's available to them both in terms of what they apply to and then if they have different offers which one they choose uh that's going to create the highest expected value for them. Again not measured just by money. Let's call it broadly their happiness right which is some combination of all the things that they're trying to achieve. And that's also done under uncertainty in the exact same way because uh it is an investment of limited resources um hoping for some outcome that on average is going to be better than other outcomes um that you might have but it's all done under uncertainty.
So basically the once you sort of understand that that what you're doing when you're picking what option to trade, right, or what stock to invest in is the exact same as choosing a job. Or by the way, like what do you what what meal are you ordering at a restaurant? That's a bet. That every single decision that you make is actually a bet because every decision is made under uncertainty. Every decision carries opportunity costs. Every decision has some expected value associated with it. whether you're explicitly calculating it or not. And uh um it's all done under uncertainty. Right?
So now that we've defined bet, we can define what a good bet is. And a good bet is what we would say is given the options that are available to you and given the resources that you have, which aren't unlimited. Given the resources that you have, a good bet is one that carries a positive expectancy. Now, the best bet would be choosing the option that has the highest expectancy compared to any other option you might choose. That's actually kind of hard to do just because of uncertainty. But a good bet is going to be one that carries positive expectancy, which is going to help you to advance toward whatever your goals are, which might be just making money. Uh, but it could be other things. Um, it could be health, it could be happiness, whatever. Um, what a good bet is not is one that wins >> because you can make a really bad bet that wins. Nor is a bad bet one that has a bad outcome or loses because you can make a very good bet and still end up with a bad outcome. So, we really have to get down this this core idea of what's the expected value? what what are the outcomes that were associated with that? What were the probability of those things occurring? What were the payoffs? Was it within your risk tolerance? Um so on and so forth. So that that's what it that's what a good bet is.
I want to go deeper into what you said about something along the lines of focusing on the process of the bet rather than the outcome.
Um can you break that down some more for me? So, let's say I made a decision and it went great, right? Why does that not imply that that was automatically a good decision?
Yeah. Because there's luck. Like, here's the thing. So, any option that you might choose has a range of outcomes that are associated with there's some distribution of outcomes that are associated with it. Which outcome you actually observe is not under your control. Now, I feel like kind of in the abstract, that's hard for people to sort of wrap their head around, but I think that if we get down to just like a very practical example, like a real world example, I think it's easier to see. Um, if you go through a green light, there are variety of things that could occur. Mostly, you proceed through the green light. That's mostly what happens. But you have like a near miss out accident. You could slip on a patch of ice that you didn't see. Your tire could blow. Someone could come from the other direction and t-bone you. If you go through a green light and someone comes through the other direction and t-bones you, it doesn't mean it was a bad decision to go through the green light.
M >> if you go through a red light and you proceed through just fine with nothing bad happening to you, I don't think that people should think that that's a good decision that they should then repeat. The reason why I think it's so obvious in that case is because there's like a well-known rule around red lights and green lights. So, we're taught, we know from a very young age watching our parents drive that um green lights are good decisions and red lights are bad decisions if you go through them, right? Stopping at red light, stop at a red light, go on a green light, slow down on a yellow. So, we know that. So, because we know that, I can make it really obvious that what happens once you've once you've decided to proceed through the green light, what happens after that isn't under your control. You can't control if you get a blowout, right? That that's just luck. You can't control what the other drivers are doing. That's just luck. So, I think everybody can see it in this particular situation. Um, I can also like give the example of like Was it a good decision for LeBron James to be tall? [laughter] Right. Like, no. That was just luck. Like, he was born to certain parents who had jeans, you know, and he inherited a set of jeans that made him very tall. So, it's it wasn't, you know, there w that was just that's 100% governed by luck. um you know if you win the lottery was it a good decision to play the numbers that you played what no like it's just luck what happened so if I if I I can come I can come up with these examples that make that really really obvious but what people need to understand is it's true for everything right like the there there when we think about why are we going to observe an outcome when we make the decision whatever option we choose It's kind of setting in place the outcomes that are available to you. Right? So if I go through a green light, there's a different set of outcomes, a distribution of outcomes than if I go through a red light. So when I go through a green light, the probability of someone hitting me coming from the other direction is much lower than if I go through a red light. So that decision is determining the probability of those two outcomes. However, it's not telling you which one you're going to observe because that's governed by luck. So, when we choose red light, green light, we've we've now determined what the probability of an of someone hitting us is from the other direction. Right? In the case of the green light, it's much lower. In the case of the red light, it's much higher. But what we haven't determined is whether that will happen. We've only determined the probability at which that will happen. Right? That's all we've done. So the reason why we can't work backwards from the outcome on one decision is because of that problem. Because even though the probability of getting in an accident is much higher if I go through a red light, it doesn't mean I'm going to get in an accident. But it's clearly a worse decision than going through a green light. So if if I can go through a green light and get in an accident and go through a red light and not get in an accident, I hope I don't work backwards from those outcomes and decide that I ought to start going through red lights and stopping at green lights. That would be pretty nutty choice. But when we work backwards from the outcome in other more nonobvious situations, that's essentially what we're deciding.
So, every single person, every single person who has ever made a trade at some point has made a terrible trade and they've won anyway, right? Poker, which is just trading. I can completely misread my opponent, get it totally wrong, like totally wrong what they could be holding. What they could be holding could not even be in my range of what I think they could be. I could just get it wrong, right? And all of a sudden, I get my money in the pot and I have an 8% chance of winning. Well, guess what? That means I'm going to win 8% of the time. And the thing that I don't want to do is win that hand and then say, "Well, that actually that was a really good decision." Because it wasn't. It was a horrible decision. And that's the thing that we need to that's why we need to stop doing that. We need to start focusing on like what was our calculation at the time? What did we know at the time? What was our forecast of what those possible outcomes were? What was the you know not just the possibilities but also the probability of those possibilities occurring? Did I have the resources? Was that a good use of my resources? Right? Like was I uh betting too much? That's a good question, right? Given what the risk was. Um how did it compare to other things that I could have chosen? And then the outcome that you actually observe is only informative in two ways. One is sometimes you learn something, you find out some information that you figure out you could have known beforehand, right? But you did like so a lot of times you learn stuff where you couldn't have known it beforehand. I don't really care about that because we're not omnisient. But sometimes you do learn something that you could have known beforehand. That's actually very useful information because you can include it in your process going forward. You can say, you know what, that's something that I could include in my process as I'm trying to calculate this stuff. And so I'm gonna I'm gonna think about that information and search for it the next time. Not in a beating yourself up way, but in a sort of eyes forward way. So that's one way that outcomes can be informative. The other ways that outcomes can be informative is you have a lot of them because then you're overcoming this volatility, right? So, uh, one coin flip, even 10 coin flips doesn't tell me very much, but a thousand does. That can tell me if the coin is 50/50 as an example, right? Like if I'm trying to figure out if it's a fair coin and I I can't weigh it, 10 coins isn't 10 flips isn't really going to help me, but a thousand flips will be very helpful for me. figuring it that out. So, you know, in something like trading, you know, or investing of any kind, what you're trying to do is across all of the things that you do, you want that basket to be a positive expectancy. Now, you're not making the exact same trade over and over again, right? Like you could be, but conditions on the ground are changing as you do that because that's going forward in time, right? So, they're not identical. Uh, and a lot of times you're trade you're you're trading different thing, you know, different things. But so can I tell that one trade was positive expectancy, one single trade? Probably not. But like over lots and lots and lots of trades, I can tell that my basket was positive expectancy because I have enough because I have enough reps. So that's the way that you do it. But like in most things that you decide in life, you actually can't create that much volume. Um to be able to to be able to actually just look at the total set of outcomes and work backwards to what your expectancy was.
So you touched on calculating the expected value based off of your forward-looking probabilities. How do you and I think that makes sense for endeavors where it's measurable or where there are clear rules as in you don't need to calculate the probabilities for certain events happening for the green versus red light because that's just common sense, right? Um but for things that are a lot more uncertain um say choosing a career in law versus software engineering, right? where there are lots of variables and you're subject to lots of variance. How do you assign those probabilities and form say an EV calculation in your head for something that's so complicated?
Yeah. Okay. So, let me just start with this. When you are making that decision somewhere, you're calculating expected value. Now, you might not be doing it explicitly, but you have to be because in order to choose one option over another, you have to it's it's often implicit, but you have to be saying that the expected value of the option I'm choosing I believe to be higher than the expected value of the off option that I'm not. And again I just want to be very clear because the word value is in expected value is that it's not just money right so we have this concept in economics and in and behavioral economics and economics in general uh in cognitive psychology called utility uh and utility is just it translates generally to happiness like what are you getting out of it right and the things that bring you happiness are are Not just money. Money is fungeible and money can buy you things that bring you happiness. It's very rare that you find someone where it's the money itself that's making them happy. By the way, um money is usually buying things that are that they think are going to make them happy, right? So, if you're somebody who wants to make a lot of money, it's because the things that you want to buy that will make you happy will cost money, right? Um, but it could be time in nature, right? That could bring you utility. That could be something that you want. Time with your children. I I talked about like fulfillment, feeling of agency. Like those things can all be put into our utility function. Okay. So, uh, so let's let's just broadly use the word happiness. So, we're trying to decide between a career in law and as a coder. Um, and we want to know which is going to bring us the most happiness, which I think is what people are deciding generally, right? I mean, I don't I don't I I don't think that's wrong. I I think that's pretty obvious. When you make that decision whether you go through this process or not you are calculating expected value because you are saying I think this career will be bring me more utility than this career. So I I just want to make sure that that makes sense just as like >> we we have to sort of settle on that as >> before we proceed with the conversation. Okay.
So if that's true, then you should be doing that explicitly >> because if we leave it implicit, that's where it's going to be subject to the most bias and error. Like we're just more likely to get stuff wrong. In the same way that if we do math in our head versus actually write it down, right? It's easier if we come up with the wrong answer to like to look at it to have other people look at it and figure out where our error was, right? Like so I mean I you know you can do some pretty complicated math in in your head but you know you're more likely to make an error, right? We all know that if you actually write it down, right? And if you go through the steps, right? And when you go through that steps, it's much easier for you to spot that you made a mistake, but it's also much much easier for you to show it to somebody else, for them to spot that you made an error. So, by writing it out, there's two things that are true. One is you're less likely to make a mistake. Two is it's easier to spot a mistake if you make one so that you can go and correct it.
>> Okay? So e whether the math is complicated or not right it's always better to write it down are you you know if you do 2 plus two equals what in your head are you mostly going to get it right sure like that's a really simple equation right and that's true with decision-m as well like low stakes easy decisions you're you're probably doesn't really matter very much uh but if you're doing calculus probably better not to do it in your head probably better to write it down but if you do do it in your head aren't you doing the same thing as when you write it down, right? Okay. So, good. I'm just I just want to make sure that our premise is correct. Okay. So, now let's take it from there. If you're doing it implicitly, you should be doing it explicitly. How do we do that explicitly? Well, you have to think about what are the things that I value because I can't I can't calculate a utility function without knowing what the things are that you value, what your goals are. Right? I don't know. So an important point in this is that what what the better decision for you because mostly other people are kind of like you. So so that that's going to be really helpful. So anything that you can think of where you could go kind of look it up to help you with that equation, you should go do that and then you can decide whether that applies to you or not. So that's that's you know, in other words, do I in do I on average think I would be happier than the average lawyer or unhappier than the average lawyer? Like that's a question that you should ask, but you shouldn't get very far away from whatever that number is, right? Because that that should be your starting assumption. The other thing you can do is go talk to people who know you well and ask them what they think, right? Which do you think is going to be more fulfilling for me? what track do you think I should be on? Why? And have them write down their reasoning as to what they're thinking about, what the upside and downside of those different careers are. Um because they know you well and they're they kind of know what your proclivities are and your tendencies are. And then you can also talk to people who are more like in a mentor situation like been there, done that, right? So you could find uh people who've gone into a law career and ask them about that. what are they happy with the choice they made? What were they thinking about when they made the choice? So on so forth. Same thing with people in coding. So that's going to help you. Then the last thing that you ought to do is do what I'd call a pre is a premortem. Imagine that I decide to become a lawyer and it's five years from now and I'm happy. Like why do I think that is? Imagine I decide to become go into computer coding. It's five years from now I'm really unhappy. you can go you why like so with computer coding for example or actually lawyer um nowadays one of the things that might be in that when you're thinking about that unhappy future is AI right so then you might think about how much then you there's another thing that goes into the equation like how much do disruption do I think there's going to be from AI how hard is it going to get you know for me if I go and I pay all this money to go to law school is it going to be hard for me to get an entry- level position if AI is writing briefs and and you know looking at case law and that kind of thing. Um and same with coding like how much coding is AI actually gonna do? Is that a career that's actually uh going to be one that's going to be uh you know get me where I want to go? Right? So you should be doing this premortem imagining these bad outcomes beforehand and kind of working back and saying what's happening why is that occurring? because that actually helps you to identify some of those outcomes that might, you know, be downside outcomes in advance that you otherwise might not consider. That is all stuff that you were probably doing intuitively but in kind of a a much less in-depth way. But because I'm asking you to do it explicitly, you're more likely to do that. Now, just to be clear, please only do this for high stakes decisions. I don't want you to do this for ordering off a menu, right? Because it does take time. But for high stakes decisions, you really should be making this stuff explicit.
That makes I want So, I want to push back just a little bit. Um because you know rightfully so we're looking at past probabilities or past outcomes and and and basing our current probabilities based on that and then forming our expected value calculation. What? How can we do that if say no one's ever walked this path before or the path that we intend to walk is extremely high in variance and so our estimate for our expected value maybe you know maybe isn't maybe isn't the best like um and I'll give the example of say starting a business starting a startup where no one's done it in the way you've done it um how would you think about that problem.
So, there's pretty much nothing that you've done that nobody's ever done completely. So, let me just start there. Other people have started businesses. They might be different than the business you're starting. They might have done it in a different way, but you're it's going to stop you from making a a really really large error. So, let's imagine that you're like, I'm going to start this business. It's a brand new product. I know it's a great product. I'm doing it in a completely different way. I think there's a 90% chance I'm going to succeed. And whi which you may in your head be thinking, but you go and you look and you say, what's the probability that startups of all types or let's imagine that it's in uh you know, let's imagine you're starting a SAS company or something like that, right? Go look at the what what what what's the probability that a a brand new SAS startup succeeds? People have done it in all sorts of different ways, right? So you assume that that's baked in and let's imagine the probability is 10%. And you have yours estimated at 90%. I hope you downgrade. Now that doesn't mean you shouldn't do it. So you may say, you know what, it's 10%, but I'm doing in this very innovative way. I think that I'm probably better than the base rate. I'm I think I'm going to be 20%. Now, you may in reality be 5%. But if you're estimating 20%, you're a lot better off than you're estimating 90%. In terms of the way you're going to manage your resources and the way you're going to think about the risk that you're taking on, um, how much runway you might raise, for example. how how much runway do I think I need in order to start this business is going to be really different in a 90% situation than a 20% situation. Right? So by looking you can always say I may not be able to get a perfect isomorph for the thing that I'm deciding to do but I can go look at a variety of different reference classes that intersect. You can think about like a ven diagram, right? that intersect with the thing I'm doing. I can go find base rates for those things. I can go find people who have done those things who might have an interesting perspective and then I can interpolate to come up with something that is going to get me like a guess a guess at mine that on average is going to be closer than the guess you would otherwise make. Okay. Now notice I'm saying it's less of an error because to your point and I think this is what you're sensing. It's not like 2 plus 2 equals 4. It's a forecast. So and it's a forecast made under uncertainty, right? We've never done the thing before. Um even if other people have done the thing before, I haven't done the thing before and I'm making a guess based on the things that I know. uh the guess is that I think that the market is going to like what I'm what I have to offer, for example, which well nobody's ever put this type of thing out into the market. So I got I have to make a guess at that, right? Um so there's all sorts of estimates that you know that are occurring, forecasts that are occurring in there and obviously it's going to be noisy, but is it going to be better than it otherwise would have? And and this is something that I really try to get people to understand is that you will not be perfect. There is not a decision that you will make in your life that is perfectly maximizing expected value. Except may maybe by accident it might happen, right? But but um but it's just because we're not omnisient because we don't have time machines. We can't get it exactly right in the sense of 2 plus 2als 4 is right. Right. can't do it. What we can do is get it better. So even though I might not get it 100%. Let's imagine that without doing this process I would have gotten it 10%. And now I can push that up to 17%. Um, imagine how that's going to acrue over time. Right now you're grinding a pretty big edge over the other version of you. who would have gotten it 10% right. So you get it 17% right. You could say, "Oh, I'm 83% off perfect." But I'm like 7% better than the other version of me. And I get to I get to those those gains acrew over time in what my outcomes are going to look like, right? And that's huge. And that's that's what we have to accept about it. And that's how we have to approach it. It It's not going to be perfect, but it's going to be a lot better than if you didn't do it otherwise. And that will acrue over time.
I love that way of thinking about it where you say it's like grinding an edge against the alternate version of yourself who wouldn't be thinking in this way. And I think that leads us to very conveniently into kind of what I wanted to ask next. Um, what makes great risktakers? What makes great bettors, decision makers, uh, poker players, traders, you know, all sorts of different careers. But what if there's what are the qualities that really make a great risk taker?
I think it's one really big quality to tell you the truth, which is to be really sanguin about the outcomes that you observe in the short run. Um, and what I mean by that is just great risktakers lean into the uncertainty. They understand that they're making decisions under risk, right? Under uncertainty and that they can't know all there is to be known and that the outcomes in the short run are going to be determined by luck. and they're okay with that. And that I think is what allows them to make more decisions under uncertainty, which is obviously taking on risk, right? And better ones. Why do I think that? Well, first of all, we know that a big factor in decision-m is loss aversion. And loss aversion specifically has to do with risk attitudes. And it comes from this problem of making you know decision-making under uncertainty. So loss aversion is part of prospect theory which is Daniel Conorman and Amsterki. Daniel Conorman won the Nobel Prize for it and it has to do with the way that we uh cognitively weight losses versus gains, right? So we ought to be calculating expected value, right? Which is what are the gains that are associated with this uh decision? What are the losses that are associated with this decision? Uh we calculate the gains by taking the gains associated with different outcomes and what the probability uh that will observe those things are. Uh and that allows us to calculate what the gross gains right associated with with a decision is. And then you do the same for the losses and then you subtract one from the other and you figure out if your expectancy is positive. All right? So that that's how we do that. Um but what happens to us cognitively is that we really pay attention to the losses that are associated, right? What's the possible downside? What could I lose here? And when we do that, it actually causes us to make decisions that are either lower expected value or might actually be negative expectancy. Um so we know for example if we go back to the startup example mostly we lose and then sometimes we win $10 billion right um now when we look across that and obviously it depends on what your situation is can you afford the risk associated with that and some people truly can't uh but but let's say that you can afford the risk associated with that and you're down as far as what you value you're down for $120 20 hour work weeks, never sleeping and the stress that's associated with that. So let's assume that that's all good for you. Uh then you ought to choose that, right? Because it is positive expectancy even though you almost always lose. Okay? Uh loss aversion would stop you from doing that because you almost always lose. So basically what loss aversion does is makes us prefer options that don't have large losses associated with them independent of expected value, right?
Or causes us to choose options where there isn't a high probability of loss. So it's both magnitude of loss, right? We don't like things that have a high like a high magnitude of loss in that set, right? That doesn't mean that the losses that are associated with it are high. It means there's some outcome in that set that has a large loss associated with it. Might be very low probability, but we see it. Or we could choose something where you just don't lose very much in it. If we choose something that we don't lose very much, you know, there's just not a high probability of any loss in there. There's probably not a lot of gain associated with it as well. That's just mathematically. though again because we now it's we're just choosing things that are very low vol um and we know traders hate it when there's no volatility because you can't make any money. Um and that's true in all decisions. Okay. And then uh if we're avoiding things that happen to have like one outcome that might be really bad, you know, might have a high loss associated with it. Again, let's assume we can afford that, right? Then if we don't choose that we might be losing out on much higher associated gains.
Okay. So, uh, that loss aversion, if you think about it, is really an uncertainty problem because I don't know what outcome I'm going to observe and I don't like the idea that there's bad things in there and that's making me choose things that don't have high expected value associated with them because I'm scared of that ambiguity. I'm scared of what I might observe because I don't know what I'm going to observe. So that's part of that embracing, right? Embracing the uncertainty, embracing that the world is probabilistic and being okay with that, right? Then we also associated with loss aversion. If you think about loss aversion, it's associated with this resulting problem. This what how are we reading into the outcome? Because the other thing that happens is when we get a bad outcome, we think, oh, we made a bad decision. And so we're afraid of getting bad outcomes because we don't want to feel like we made a bad decision. If you've leaned into the uncertainty in the first place, then you understand like, look, I went through a green light and something bad happened. And that's okay. So you have to really embrace that. All of that ends up causing us to be unwilling to take risk because we don't we don't like the volatility. We don't like the unknown. And then the other thing is that because of this problem sort of what you've been hinting at, right? Which is but what about things that are really complicated where you don't actually know what the expected value is, right? And I'm saying to you but you're 7% better than the other version of you, right? That's that's accepting the uncertainty is that it causes us to be paralyzed because we don't want to make a decision unless we know what that equation is. We want to know that it's not just 2 plus 2= 4, but that it's 2 plus 2, right? And what we're really in is like a plus b equals c. And A is anywhere between 2 and 225 and B is anywhere between, right? And maybe you don't even know if there's an addition sign in there, right? And people don't want to act under circumstances where that's what the equation looks like. They want to know what those variables are. They want to know what the answer is before they start because they don't ex they're not comfortable. They're not comfortable in the uncertainty and they're not comfortable acting in the uncertainty. So in order to be a great risk taker, you have to embrace that. You have to be comfortable with it.
I love that. And I think you addressed my earlier concern of, you know, analyzing potential decisions where you where they're extremely complicated. And you're saying that it's much better to at least have an estimate than to just say nothing at all. Right.
Well, for a variety of reasons. One is that that you're more likely to make a better estimate because again there is no such thing as nothing at all because it's in there somewhere. You're choosing option A over option B. How? If I said to somebody, why'd you choose option A over option B? They didn't go through this process. >> Why' you choose option A over option B? Well, I thought option A was better. >> Okay. Well, what does better mean? Better means higher expected value. Okay. So, so when you say like doing it versus not doing it, there's no such thing as not doing it. There's no such thing as not doing the calculation >> because you you did it. >> Do you see what I'm saying? >> Yeah. >> It happened. >> Yeah. >> Now, it feels better because nobody's going to mark it wrong, right? [laughter] >> But it's not better because there's more likely to be error in there. You did math in your head, right? math in your head isn't as good as as math on paper. Like what if [snorts] what if you're a trader and your quant comes to you and says, "I think you should make this trade." And you say, "Why?" And they they go, "I I just I I've got I've got a nose for this thing. I just think this is better." You be like, "What?" Now look, the quant knows that sometimes their work isn't correct, right? the math is correct in terms of the equations, right? But sometimes the inputs aren't right. They understand that, right? But you ask them to do the quant work anyway because then all of that is explicit, right? And you can look at what the assumptions are. And sometimes you can look at that and say, you know what, I think your assumptions are wrong that you've put into the equation. Thank god they're putting it on paper. You would never accept work from a quant who was just like I think you should do this instead of this but because I did it in my head and you'd be like what? But we accept that all the time from people in other types of decision- making right and because with the quant you're like wait but you did math somewhere like could you just put it down on paper like tell me what it is that you were thinking about. So that that's the first thing is putting it down on paper makes it more accurate. The second thing and this is true in quant work as well is that on the look back because again this is decision-m under uncertainty. Every decision you've ever made you're going to learn something new after the fact. It's just true because you're not omnisient. And also you didn't know what the outcome was going to be. So at minimum you're going to learn what outcome you observe. But you're also often going to learn new stuff. Okay. We can go back now and say, did were there things in there that we didn't actually get right? As I was thinking about this, right? So, you might learn new information. For example, like let's imagine that you're making a trade and it's because of some forecast you've made about interest rate movement. And then stuff happens in the world, right? And I'm sure this has happened to everybody where you go, oh, my estimate of the probability of interest rates going up was nutty. [laughter] Now, you can go back and you can say, could I have known that, right? Could I have known that? And a lot of times, the answer is going to be no. I I I really couldn't have known that. But then you should still ask, but now that I now that this has been revealed to me, could I could I actually know this going forward? And sometimes the answer will be no that was random. Sometimes the answer will be yes, you can include that. Sometimes your answer will be yes, I could have known that beforehand and I didn't. Great. Then again, then you can then then you can include that going forward. Um, but it's the fact that you've got the work there that allows you to see like, oh, I really didn't include that in my calculate. Like, as I was trying to think about that, right? Like here here's an example, right? Up until recently, there has never ever been a threat, I don't think, of a Fed chair being fired. I don't think there I don't I I might be wrong about that, but I don't think so. So, when someone is trying to calculate what the probability of interest rates moving in a certain direction are, they're probably not including Fed chair gets fired, right? And they're looking at that particular Fed chair and they have a lot of they've got a model of how that person thinks, right? They know what their past behavior is. they kind of know what their thesis is about the relationship between [snorts] um interest rates and inflation. Uh they know whether they're kind of like a hawk or a dove when it comes to that stuff like how aggressive are they on the this and you know so on so forth. So, so they can probably make some very good forecasts about given what the current economic uh situation is, how do I think that Fed chair is going to behave? And then all of a sudden, the Fed chair gets fired right there. You're just like, could I have known that? And it's like, well, look, it's reasonable for me to look historically and say, this is something that even the Supreme Court has said you you you shouldn't touch. And no, I couldn't have known, right? Could I know that going forward? Okay. Yes, that that I could think about going forward, but it was reasonable for me not to include that. So, I I mean, I think that that's how we have to actually think about that stuff when we do that. And it's okay that you miss stuff because it's better for you to at least think about that explicitly because again, not only will will your estimate at the time be better, but we can actually take a look back and and improve our decision-m going forward in a better way based on the world that we actually end up observing.
Where does intuition fit into the process of decision- making?
Yeah. So um you know people talk about the gut feel a lot and there there let me sort of divide this into two categories right? Uh the first is kind of how good is your gut and the answer is sometimes good sometimes bad. Like it's certainly the case that sometimes whatever your in your your intuition is going to tell you is going to be pretty darn good, but it's also the case that sometimes what your intuition is going to tell you is going to be pretty darn bad. And and part of the reason for that is that your gut is uh subject to the most cognitive bias, right? So, if we think about um as an example, let's imagine uh that we're looking at a stock that's trading at 40 and we do a bunch of analysis on it and we decide it's not a buy, right? Uh if you own that stock, if you bought that stock at 50, your gut and and you think about that trade at 40, right? your gut is more likely to push you toward your intuition being to hold that stock. But we know if it's not a buy, you should sell it less transaction cost, right? We we know that, right? Because there's huge opportunity cost to that capital being in something that you don't think is a buy because if you don't think it's a buy, you're you're saying it's negative expectancy or the expectancy isn't high enough for what my uh what my portfolio wants, right? So, um, confirmation bias, right? Like your gut is going to lead you to things that confirm what you already believe because you're not actually able to examine it. So, sometimes it's going to be pretty good, sometimes it can be pretty bad. Um, what that means is that when you make a decision by gut feel, the probability that you're introducing error into that decision is higher. That's just true. Now, that doesn't mean that you shouldn't make any decisions that way. It means that you should think about the decision you're making and how tolerant you are of introducing error into the decision because sometimes we're very
tolerant of error, and sometimes we're not so tolerant. And that has to do with, um, two ideas. Idea number one is, uh, what is the long-term impact of observing a bad outcome, not what the short-term impact is, because the short-term impact, we're all losers, right? And so we know the short-term impact of ordering something at a restaurant and getting a dish that you don't like is like devastating. Like, I'm so sad, my life has been ruined. But if I talk to you a week later, it's your life was not ruined. You were just sad in that moment, right?
So, I think about this as like the renting versus buying. If I rent an apartment and I don't like the neighborhood and it's not great, it, it's not that big a deal, right? Generally, particularly if I can find a month-to-month lease or a six-month lease or something like that, it's not going to have a big impact that I have to live there for a little bit. Um, dating versus marrying, hiring an intern versus hiring a co-COO, you know, there's, there's all sorts of things. Ordering something off a menu, like, who cares, right? Like, okay, you didn't like your dish that much, whatever. So, I, I think that when the impact is lower, the long-term impact of observing a bad outcome is low. Um, go, I don't care. Go with your gut. Like, if you introduce some error into that decision, it, it just doesn't really matter very much.
The second thing is, um, uh, the more equitable the decision is, the, uh, the more that you should be tolerant of error. And this is actually when we go back to your question about what makes a good risk-taker. It's understanding optionality is in there, right? Because we can take more risk when, a, there's more optionality, and, b, we're willing to exercise it. Both things have to be true. So we could be in a situation where there's a lot of optionality and we don't exercise it. In that case, right, is there, it's like a tree falling in the forest, right? Like, if we're not willing to exercise an option, then, then the option, the option might as well not exist. But, um, you know, so, uh, you know, this is this idea of like, you know, Silicon Valley came up with like minimum viable product, it's kind of taking this into all this into account. Like, I can move faster if I'm doing smaller things that will have less impact. In other words, I'm not releasing product to my whole customer base, right? I'm making a small change to some small number of people that reduces impact, and now it makes it easier for me to pull it back if it doesn't work, right? I, I, I just. So, so this is actually very important conceptually. Um, gut, I don't care. [laughter] Whatever. Now, the reason why you don't want to go with your gut and options trading, even though there's optionality associated, was because impact is high, right? That's why those two things are separate. So we need to think about both. It's not just that there's optionality, right? Um, we have to, we would have to have low impact along with. Does that make sense? Okay.
So that's the first thing about gut is like, understand when it's okay for you to make errors, because when it's okay for there to be error in there, like, I don't care, just whatever, gut away. But then the, the other thing that you, you had mentioned is, but what happens when, like, you're doing these calculations and then something doesn't feel right? So that would be a gut thing, right? Your, because your intuition is telling you something doesn't feel right. I absolutely think you should pay attention to that. But what I want you to do is figure out what doesn't feel right. Can you articulate what doesn't feel right about this? That's the thing that I want you to try to figure out. Ask other people, like, "Can you look at this and tell me what this looks like to you?" Or ask them to make the same decision independently and see if they're spotting something, because generally, I think that [snorts] doesn't feel right could be coming from a few places, right? One is you could have actually missed something, right? Like, and, and you just haven't identified, like, there's some, some, there's some input that you haven't gotten explicitly. Your gut is sort of sensing that you're missing it. In which case, that you would want to change, you may want to change your mind because it's saying like, "No, go look for more information." Like, I think you're missing something here. And maybe you really are missing something.
Sometimes it doesn't feel right because of bias, because the answer is not what you wanted to hear. In that case, you ought to make the decision anyway, right? So that, that's why I want you to listen to it, because both things could be true. Your intuition could be saying you've missed something here. You've made an error, and you should explore that because you would want to spot that error. But your intuition could also just be saying, I don't like this answer. The answer's right, but it doesn't feel good to me because it's not the thing I wanted. Um, you know, in that case, you should make the decision. You should make the decision anyway, which is why you have to figure out what it is that your gut is telling you. Sometimes you won't be able to, and that, that's okay. Like, then, you know, on balance, I would probably lean on other people in terms of helping me sort that. Uh, sometimes you just got to shrug your shoulders and say, I, I got to, I'm going to make a guess at what this is, but like, you, you should do, I do want you to pay attention to that because very often it is telling you that you missed something.
You know, I, I'll tell you what it is. Like, this is, this is where I think [snorts] this might become clear. So people very often ask me like, "Is this the way you think about marriage?" And I'm like, well, of course I think about marriage that way. Like, if I choose to marry somebody, I think that of all the people that I imagine I'm going to meet within the time frame that I have, that this person is most likely to, well, they're going to bring me the most utility, right? That, that I'm going to feel the most fulfilled in my marriage, whatever my, whatever it is that I think is going to make a happy marriage. Utility in a happiness sense. Utility in a fulfilling happiness. Exactly. >> Um, that this person is the best match for me given what I want. Now, you know, it's important to understand that people grow, right? And I might make that choice and then 10 years later, that might not be the case anymore. They might change, I might change, I might discover that what I wanted isn't what I thought I wanted. You know, those kinds of things. So, so it does change over time, but at the time, like, that's what I'm doing. I'm like, I think this person is the best for me out of anybody. Um, and I, people really bristle at that. But then I say, plus love, right? Like, plus love. So it's true that you have to think about like, do we align on how we want to raise kids? Do we want to have kids? Where do we want to live? How career-focused are we? Like, all those things that are things that are important to you and that you value, but someone could check all those boxes, but you don't love them. And I think that's sort of in the category of what you're saying, right? Is that there's this thing that has to do with how does this feel okay for me? And, but the thing that I want you to ask is, is it a thing like plus love, or is it, but this means I have to change my mind and the belief that I had was wrong. I thought this would be the good thing, but the right. So you just have to separate those things from each other.
>> How many kids do you have, Annie?
>> I have four plus two steps.
>> How do you raise them or what advice do you give them to, I guess, from a young age? Are you teaching them how to make great decisions? Is that what you're doing as a mom? Are you like, "Okay, here's the EV," you know, like, how do you, how do you do it? How do you do it? I'm curious.
Um, so there, I, I think that there's a lot of tools that make for great decision makers, and I think that I, I just naturally do those for them. So one of the best tools that I think you can teach children is mental time travel as a skill. And the reason is that.
>> Can you break down what mental time travel is?
>> Yeah. So remember I talked about, uh, premortems, right? So imagine it's five years from now and you're really unhappy in your job. Uh, that's a form of mental time travel. So human beings have this ability to imagine the future, and in fact, a pretty distant future. Um, and imagine themselves in that future. And it's actually a really, really important skill [snorts] because when we're in the decision, or when we're in the moment of getting bad chicken at the restaurant, it, it has a very outsized effect on us. And we actually tend to make our worst decisions in those moments when we're facing it down. If we can imagine the future, right? Like, how is this going to feel to me in the future? Or if I do this thing now, is me two weeks from now going to be happy that I made this decision? Right? This is like a great way to, to actually make better decisions because notice what have I been talking about. Imagine what the possible outcomes are. What are the probabilities of those things? You know, okay, well, that's time travel, right? So, if you can start to get yourself into the future more, you're going to tend to make better decisions. You're going to tend to have a better, what we call discount function. Meaning our tendency as human beings is to discount, take a discount now. Uh, so in other words, like, do something that's kind of like bad for us long run because it feels better to us now. I'm going to eat a cupcake now even though it's bad for me in the long run. Right? [snorts] I'd rather sit on the couch and binge a show than go to the gym, even though future me wants me to go to the gym. Okay. So, uh, this is just a very common phenomenon. It's called temporal discounting. But if I can say, if I can stop and think and say, but if I binge watch this show and do and go to the gym, when I wake up tomorrow, am I going to be happy I made that choice or sad? Well, tomorrow me is going to be sad. I'm going to beat myself up for not having gone to the gym. So, if I can talk to tomorrow me, then I'm more likely today me is more likely to go to the gym because I've had a conversation with tomorrow me. This is a really good tool for kids. And it's a good tool not just to get them to start thinking about like, you know, if you play video games instead of studying, like, how, how do you think you're going to feel when you're taking that test, right? But also, this was a tool that I used with them all the time. Was I would be grounding them for something, you know, because teenagers and they would be really, really upset and they would be arguing with me. And this is what I would say to them. Imagine how awesome this is going to be when you're like 40 and it's Thanksgiving and you're at the table with your family and you get to tell your kids about how grandma grounded you. And then they would start laughing because obviously that's amazing. And what brings me great joy is my children who are now young adults. I've heard them now say that, right? Like, something's happening and they're like, "This is going to be a great story in the future." Right? Because it allows them to see what is the longer-term utility of this thing that has occurred as opposed to in the moment it feeling like disutility. Right? So when they're being punished, it feels like disutility. It feels like a bad thing. But I point out to them, you get to tell this story for the rest of your life, and then it, it actually changes their viewpoint of whether that thing is good or bad. Right? So, so I think this is it. Those kinds of things when you're wrapping that in gets them to start thinking in expected value. It gets them to start thinking about utility in a way where I'm not decision scientist mom. [laughter] You know what I mean?
>> Here's the EV, kids. You know.
>> Right? And then, you know, and then with my kids, it was a lot of conversation about luck, right? So, uh, we look, I worked very hard in my life. I want to be very clear about that. But I was also very lucky. Um, uh, just, I mean, in the simple sense, right? Like, I was born at a time when the most probable outcome for a woman wasn't secretary mom, right? Like, a little like, a couple decades before I was born. Secretary, wife, mother, and very rarely something else. A century before I was born. Whoa. Like, that gets really different for a woman, right? So, I, I was born at a time when there was a lot of opportunity that was available to me. Um, you know, I was able to go to graduate school, like women were getting it, it was still a little rare, but women were getting PhDs. Uh, not only did I start playing poker, but eight years into my poker career, it was on television. I have no, I didn't know that was going to happen, right? And obviously, that really changed my trajectory and the, and the types of the types of opportunities that were available to me. The distribution of outcomes that were available to me was very different once poker was on television versus not. Um, so there was just a lot of luck involved. Uh, and I just had a lot of discussions with my kids about, look, I get, I work hard, but like, you need to understand that there's a lot of luck in the way that things have turned out for me, for you. And there are lots and lots of people in the world who work much harder than I do, who, who have three jobs, and they're no less deserving of what we have than we are because there's just an influence of luck on what happens to you. Um, and I would say to them, that doesn't discount that you have to intersect luck, which is just sort of like, through no fault of my own, what opportunities are available to me, right? And you have to intersect that with your hard work and your decision-making and that kind of thing, because that's going to determine, it will change the probabilities of different things occurring, right? You can push yourself out to the right tail. And I just made that like, really, really clear to them. I think, um, in a way that I think has, has really helped them. So, you know, I don't think that I need to sit down and like map out decision trees with my kids, right? Because that's happening naturally in the conversation in terms of the way that we're thinking about stuff. When they're asking me for advice, I'm walking them through process. I'm having them do mental time travel. I'm having them think about luck, right? And I, and I think that that in the end is really, like, you're, that's going to make them good decision makers.
>> If there's one thing that you think the person watching this should take from our conversation, guess what is the single most important thing in all that you said? I'm thinking about this conversation because there's a lot of stuff. There's things, you know, I have different answers depending on what the focus of the conversation is. Um, I think the single most important thing from this conversation is what we started with, that every decision is about, right? Which really means every decision, uh, is a forecast. Contains involves a forecast of the future. Thinking about what you're going to invest your limited resources into and understanding that because it's a, a bet, that all that you've done is invested in a range of possibilities that have a certain probability of occurring. In other words, you must lean into that uncertainty and embrace it. And I honestly think that you'll be happier if you do that. I really do. Uh, because you're neither going to live in a world where you think the world is just happening to you, which I think is the road to unhappiness, where you don't feel like you're an agent, right? Where it's just like, stuff is happening. Nor are you going to be in a world where you think you have so much control over the outcome that you're going to be disappointed a lot, right? Like, you're going to be so surprised by the world. You're just going to live in constant surprise and disappointment. And instead, you're going to live in this middle place, right? Which is like, the, it's probabilistic, but the option I choose, and if I really put thought into that and think about it that way, the option I choose is going to determine what the probability is that I get in an accident, and that I, I am an agent of that choice in this probabilistic world. I think that's the most important thing for people to really internalize and, and start to build into the way that they approach decisions. And I think I, it's not just that I think you're going to be more successful. Like, I really do think, like, you're going to be happier. You're going to be more emotionally regulated. Um, I think it's a road to being calmer. You know, separate from anything else.
>> I love that. I think that's a great place to to wrap up. Thank you so much for coming on Odds on Open. All the best.
>> Well, thank you.