The Value Perspective Podcast episode – with Annie Duke

Chapter headings for Annie Duke on The Value Perspective Podcast

Please click on the link below to jump straight to a chapter

* Annie Duke, welcome (back) ...

* Why is it so difficult to quit?

* ‘Grit’ or ‘quit’ is the same decision

* Why quitting is such a powerful tool

* Thinking in terms of ‘expected value’

* ‘Kill criteria’ and ‘pre-commitment contracts’

* Could anything help Putin quit Ukraine?

* ‘The hardest thing to walk away from is who you are’

* Three book recommendations

JTR: Annie Duke, welcome back to The Value Perspective podcast. It is a pleasure to have you back. How are you?

AD: I’m good. How are you? I mean, I broke my wrist but other than that!

JTR: Oh no – what happened?

AD: I was playing tennis and I went for a very high ball-poach. I hit a winner and then fell on the ground and broke my wrist. So I’m out until ... I still have about three more weeks to go.

JTR: But if you hit a winner, you must have mixed feelings! Did you feel confused about what to feel about the whole outcome of the situation?

AD: Well, I’m proud I hit a winner! But I’m sad I’m out for five weeks. I would go for the ball again – I mean, it was a perfectly good poach – I just lost my balance, that’s all. But it’s fine – it’s not that bad. I don’t need surgery so that’s the really important thing to know.

JTR: Well, we are really sorry that happened to you. Now, regular listeners to this podcast will need no introduction to you ... actually, I am not sure if you know this but this podcast has now been around for three years and the reason it came into existence was because of you.

AD: No – I definitely don’t know that.

JTR: Well, I don’t know if you remember this but back in – I think it was April of 2019 – you very kindly gave us an hour of your time to discuss your book Thinking in Bets and we recorded the interview so we could write some pieces for The Value Perspective blog. And everything you said made so much sense to us, we decided we wanted to explore in more depth how to make better decisions under uncertainty. Three years and something like 60 episodes later, here we are and now, on your third appearance with us, you are here to discuss your latest book and the powerful lessons we can take from it in the context of decision-making. So, first, thank you very much! And then, for those who may not be familiar with you, please could you provide us with a little bit of an introduction?

AD: Sure. That’s a nice thing to know – I did not know that and I’m happy you have had 60 episodes, in small part because of me. So just real quick – I started off as an academic, doing five years’ worth of PhD work in cognitive science at the University of Pennsylvania and then, right at the end, I got sick and had to take some time off. While I was taking a year off from graduate school, I found my way to poker as a way to make money and it turned out I had a knack for it. I was pretty good at it and so I ended up not going back to graduate school and not finishing – or ‘ABD’, as they say. ‘All But Doctored’

I played poker for 18 years, retiring in 2012, and was lucky enough to win some World Championships along the way so that was fun. But in 2002, I started getting asked to speak to groups about ... well, the first group asked me to speak about how poker might inform their thinking about risk. I talked a little bit around that topic – in the sense that I was talking about how poker can inform the way that whether you’re in the losses or in the gains will affect your risk attitudes, which I thought was a more interesting topic for them.

They seemed to like it and, in that moment, I remembered that I really love teaching. I had a very academic side of me, which I liked, and I decided I wanted to dive deeper into that conversation between poker and cognitive science. So I started giving more talks and getting referred by people and eventually started getting asked to do consulting. And I also started teaching poker, by the way, because I just really wanted to get back into teaching.

In 2012, I retired from poker and started doing speaking and consulting somewhat full-time but also started a non-profit called the Alliance for Decision Education, and then decided I really wanted to write this book. Thinking in Bets, which I did. I followed that up with How to Decide and then my latest book, Quit, and then made my way back to the University of Pennsylvania where I now teach effective decision-making in the Executive Ed programme there.

And, in the fall of 2022, just recently, I re enrolled as a graduate student and I’m now finishing my PhD with Philip Tetlock and Barbara Mellers. For those who don’t know, Phil Tetlock is the author of Superforecasting and I had done a bunch of research with him during the pandemic, because I’m active in a couple of labs there. That resulted in four pretty largescale studies, the results looked pretty good, it was training people to become better forecasters in a variety of ways – and he said, you know, you could just write these up and make these a dissertation. So here I am and I’m hopefully going to be ‘defending’ [my thesis] in the spring.

JTR: Oh wow – that’s fantastic. Many congratulations!

AD: It is weird – but that’s a good lesson about quitting by the way: the things you quit you can often go back to.

Why is it so difficult to quit?

JTR: That is very true. Now, you are a good friend of another guest of ours, Ted Seides, I believe. He has a fantastic podcast and he did this great episode on negotiation with Professor Daylian Cain. And in that specific instance, Cain was explaining how his biggest failure as a teacher so far had been to make his students understand and act on the importance of being willing to walk away from any negotiation. He said he had tried many different approaches – even threatening to fail the students – and not even that had worked. So, by the time that episode was recorded, he had given up trying and was rethinking his approach – so let’s start with why is quitting so difficult?

AD: Well, I have some news for him! The way we kind of judge ourselves is in comparison to the goals we have – or to some sort of benchmark. So, in the simplest sense, if you bought a stock at 50 and it is trading at 40, the benchmark is 50 and you’re short of it, right? You’re in the losses. But there are other ways we benchmark ourselves as well – like, we can think about ‘in the losses’.

In that particular case, you could have a physical ledger – you know, your actual P&L – and what you would see is that, cognitively, you would be ‘in the losses – $10’ and also, on your P&L, you would be ‘in the losses – $10’ But that feeling of being in the losses is a cognitive phenomenon that does not always line up with your P&L. I’ll give you another simple example of that: if I buy a stock at 50 and it goes up to 75 and then down to 60, I’m cognitively in the losses – even if, on the P&L, I’m in the gains. Does that makes sense?

Then we can also have goals, right? So if we’re going into a negotiation, we have some sort of goal for how that is going to turn out. That’s our benchmark and, anything short of that, now we’re in the losses. So this idea of being ‘in the losses’ becomes really important for why we won’t quit because, whenever we open up a mental account for something, when we get in the losses, it’s only if you close the account – in other words, if you quit – that you cannot actually recover those losses. You can’t wipe that loss off the books – again, the cognitive books, right?

So you can think about this moment of quitting as ‘failing’ – but it follows, if I quit, I have failed. As long as you keep going, there is some chance you can recover the cause, right? So if you want an umbrella idea, that would broadly be what is going on. Now, it turns out there are a lot of cognitive biases against quitting that kind of line up to fill that space – many of which I’m sure a lot of your listeners have heard about.

Like, you can see how a sunk cost fits in with what I just said, right? You have sunk time, effort and money into something and it is only when you walk away that you have this feeling of having wasted that time or wasted that money or maybe of not being able to recover that effort you put in. That is a fallacy because what we should care about is – is it worth it to put the next bit of effort or the next bit of money into it going forward? And what you have already sunk into it actually should not enter into the equation at all. We can talk about that more.

And then there is endowment, issues of identity, status quo bias, omission/commission bias and so on – and then this broad problem of escalation of commitment. So essentially, when you look at that, the deck is really stacked against us when it comes to quitting. And if I had to put it into a nice little bucket, it is – when it is correct to quit, it is still uncertain whether you would actually fail if you continued or whether you might succeed. As with all decisions, right, it’s a forecast of the future. So that is problem number one – and then problem number two is we don’t like to close mental accounts and losses. And there are all sorts of ways we get into that cognitive state of being in the losses – and then we won’t walk away.

‘Grit’ or ‘quit’ is the same decision

JTR: Very early in your book, you talk about ‘grit’ as a concept that goes in opposition to quitting ...

AD: Just to interrupt, it’s not in opposition to quitting – the problem is, people think it’s in opposition to quitting. It’s the same decision – it is just a stick-or-quit decision any time you are pondering something – but we think about them in opposition, which I think is part of the problem. So anyway, now, continue!

JTR: Thank you very much for that clarification. But something you point out very early in the book is what people who have been winners in the past always seem to have in common is perseverance. They persevere and they never quit so how do you reconcile that with this mental mode of ‘Never give up’?

AD: As I said, ‘grit’ or ‘quit’ are the same decision but we don’t think about them that way – we think about them as opposing forces. One is seen as a virtue; one as a vice. I don’t need to tell you which one is seen as a virtue – we know. Because if I called you a ‘quitter’, I think you would think I was calling you a ‘loser’. So, you know, perseverance is usually seen as an act of heroism, right? That is kind of how we think about it –when we talk about building character, we talk about ‘grit’.

So how do we think about why these two things need to live together? Let’s first of all go back to your point, which is – it is true that anybody who has succeeded at something has also stuck to it? It is true but it is also solipsistic – it is just recursive. So anyone who has succeeded has stuck to something. The problem is that is true in retrospect and we turn it into something we think is true prospectively – that if you stick to things you will succeed. But that is not true.

What is true ... and by the way, I highly recommend you go ahead and read Angela Duckworth’s book Grit: The Power of Passion and Perseverance. It is a great book – and she would agree with this. It’s not that if you stick to things you will succeed, it’s that if you stick to things that are worthwhile, you’ll succeed – even when they’re hard. Even when you face obstacles, if you stick to things that are worthwhile, you shall succeed. But what will make you not succeed is sticking to things that aren’t worthwhile. In fact, sometimes that really puts your life in danger, right? When people continue up Everest in conditions where they should not be, they are very gritty but that is not the road to success – that’s the road to dying on top of that mountain, right?

So the idea is this – and we can think about it from the standpoint of an investment portfolio – you want to sample a whole bunch of stuff and then you want to concentrate your capital on the things that are working and quit all the rest. That is the actual road to success. When we think about quitting, we think it’s going to slow our progress – it’s going to impede our progress to be being successful. But that is only true if you quit things that are worthwhile. What will actually impede your progress is sticking to things that aren’t worthwhile – because, when you’re sticking to things that aren’t worthwhile, that is time and energy and money and attention that you can’t put toward things that are worthwhile.

If you are on a path where you are not gaining ground toward your goals – or not enough in comparison to other things you might be doing – then not switching, not quitting and going and doing something else, is actually going to stop you from succeeding. So if you want to get to where you want to go faster, you have to be both ‘quitty’ and ‘gritty’. It is about figuring out what is worthwhile and what isn’t. And then you want to take these two things together, which really are the same decision, and stick to the stuff that’s worth sticking to – and not only quit everything else, but quit them as quickly as possible. Find out fast. Get it off your books. Go do something better.

JTR: Another thing you mentioned in your book – and it was touched on in that Ted Seides podcast too – is that successful people quit a lot. They do not keep to one specific path all the way.

AD: Yes – and we can think about this across the board, right? Let’s take salespeople – the most successful salespeople qualify out their leads quickly. They actually want to do that because they are concentrating their grittiness – because salespeople are very gritty – they are concentrating their grittiness on the leads that matter ... on the ones that are actually going to close. And they’re trying to qualify everything out early besides that. So you want to have this very fat tail of distribution, quit fast and then stick to everything else pretty long.

If you think about investors, the best investors – and I’m not talking about retail investors who probably should be indexing the market and just holding; I’m talking about people whose opinion is a good opinion about the market – the most successful investors are putting things in their portfolios and then they’re selling the stuff that goes against their thesis and they are trying to do that pretty quickly. And they’re holding or concentrating capital on the stuff that is working.

A good example of that would be somebody who is, say, an early-stage venture investor – they’re buying lots of options, they’re investing in a lot of companies early and then what are they doing? They’re figuring out which of these companies am I supposed to buy up at A, buy more at B, buy more at C? And that’s not every single company. And one of the things we need to understand is that is a form of quitting, right? In a situation where once something is in your portfolio, you can’t let it go, what is the ‘quit’? The quit is I am not concentrating capital into this set of companies, which is going to be most of them. And I am concentrating capital – that’s the ‘grit’ – over into this set of companies. And they’re trying to do that really, really well because the initial investment is made under high uncertainty but, as they start to collect that new information, they’re going to quit fast, right?

Think about Silicon Valley and the idea of ‘agile software development’ – it is a really good implementation of the idea of quitting, right? We’re going to try a bunch of stuff; we’re going to see whether customers like it; and, if they don’t, we’re going to stop. And if they do, we’re going to continue. They are creating a minimum viable product so they don’t put very much effort into something. They’re trying to figure out, what’s the minimum we can do to get the information we need to know whether we should quit this, or whether we should start to concentrate our capital on this? Capital being time, attention, human resources of the coder and so forth. So that is also a very good example.

It’s really what scientists do too – you’re exploring lines of research; it’s not fruitful; and you say, I’m abandoning that and I’m going to try this other thing. So it’s true whether you’re in product management, product development, sales – really, if you’re a chief strategy officer, you have to be testing things out all the time. If you’re in marketing, that’s what an ‘AB’ test is – it’s just the plan to quit. I have two ideas; I’m going to test both of them; I’m going to quit the one that doesn’t work out; and maybe I’m going to quit both of them because maybe neither will work out. But I know there’s quitting built into my plan there – and what you don’t want to do is pursue one that actually was kind of a dud, right? You want to quit. So I think this goes across everything you do – and, by the way, perhaps this is most obvious in something like poker, where folding a hand is quitting. Elite players fold way more than other people – and it’s a big piece of their success.

Why quitting is such a powerful tool

JTR: In Quit, you have written one of the most powerful lines in all of your books – in my humble opinion!

AD: OK, Juan!

JTR: You write: “If I had to skill somebody up to get them to be a better decision-maker, quitting is the primary skill I would choose – because the option to quit is what allows you to react to the changing landscape.” Please could you elaborate on why quitting is such a powerful tool?

AD: It’s partly, by the way, because of the professor who said he couldn’t get his students to quit, which I’m not surprised about. So here’s the problem we have as decision makers – and let’s go back to Thinking in Bets for a second, which is why this is a good follow-up to that book. Thinking in Bets was all about the problem we have making decisions to start things. What happens when we’re making a decision to start things – whether it’s hiring an employee, taking a job, getting married, deciding if we want to create a start-up or invest in a start-up or if we want to, you know, play a musical instrument?

It doesn’t matter: whatever decision I make is going to be made under conditions of uncertainty – sometimes more, sometimes less – and that uncertainty is going to come from two sources. One source is going to be plain old luck. I can make a decision that is going to work out 80% of the time and, by definition, that means it’s not going to work out 20% of the time – just by definition – and I don’t have any control over when I’m going to observe that 20%. So that’s just luck – the pandemic was luck, right? It affected a whole bunch of decisions people had made in terms of the way they turned out and obviously – literally – nobody had any control over that. It just happened. Right?

So now we can think about the second source of uncertainty and that is the quality of the information you have that is being inputted into the decision. And the quality of that information is affected by two things. One is just hidden information – that the information we have is incomplete and we know very little in comparison to others elements we do know. And then, in terms of our own information, it’s riddled by cognitive bias, which we also know. So things like confirmation bias, for example – as you’re trying to think about how to construct the decision, that’s also going to affect the quality of the information you’re inputting.

So you have these two problems, right? But we decide things nonetheless – because we have to. There is pretty much no decision you’ll ever make where you have perfect information. And Thinking in Bets was all about embracing that and saying, OK, that’s fine because – and here’s the reason why quitting is so important – if you’re making decisions under conditions of uncertainty, what that means is, after you start something, information discovery will occurs. I’m going to find out there’s a pandemic – as an example. I’m going to have that feeling – which I’m sure you have had, Juan – of ‘I wish I knew then what I know now’. That is that feeling of, oh, I discovered new stuff after I started this.

So the question is, if you’re going to discover a whole bunch of new information after you make a decision, how do you deal with that? Well, you can quit – when that information is bad news. And in fact, the option to quit – that option we have all been given – is what makes decision-making under uncertainty something we can even do, right? We can make decisions relatively quickly – in comparison to our lack of knowledge and how much luck is involved in the outcome – because, after we find out new information, and that information is bad news, we have the option to quit.

So this is the most important skill, I think, people need to have – when you are getting signals from the world after starting something, you have to be really good at recognising when it’s the appropriate time to quit. And that is a forecasting problem – it is looking into the future, just as the starting decision is a forecasting problem. I’m trying to figure out, among these options, which thing do I forecast to be the highest expected value that will cause me to gain the most ground toward my goals, given what it’s going to cost me? That is what I’m thinking when I start things and that is what I also need to think about when I stop things.

So that will also be a forecast into the future. And that’s where the problem is – the quitting decision is itself made under uncertainty. In other words, when it is correct to quit, it will not be 100% that things will work out poorly. If it were 100% – and we could apply that to climbing Everest – you’ll be dead. So you should be making that decision long before it is 100%. In fact, you want to make the decision when the expected value goes negative or gets low enough in comparison to other things you might switch to.

And what that means is it is just a really hard skill because you’re asking people to quit when it’s not 100% or it’s not obvious people ought to quit. And that’s what the professor is finding out with his negotiation problem – that people are going to bang up against that until they know it’s not going to work. In other words, absolute failure must have occurred. And in fact, in order to avoid that, they might take terms that they shouldn’t take, just because they want to know – could I have turned it around? Could I have made it work or not? So it is just an incredibly important skill to develop, given the conditions under which we have to start things.

Thinking in terms of ‘expected value’

JTR: As I was reading the book, I kept nodding at all the things you were saying – and thinking how difficult it is to make an active decision to quit. And one of the most important variables that needs to be considered, which you alluded to in your previous answer, is this concept of ‘expected value’. I was discussing this with my colleague, Andy Evans, who said, well, why don’t we ask Annie the following question and see what she thinks? Are there any weaknesses to thinking purely in expected value terms? What he had in mind was when time is a factor and you are no longer judging a decision on a stand-alone basis – instead, there’s a continuum, things are interlinked and then you might find yourself in a situation where by quitting, you are quitting all the time and that makes no sense.

AD: The answer is no – because you take that stuff into account. Expected value is looking at what the gains are over time versus what the losses are over time. So that can be things like time – and then there is something called ‘implied equity’. When you are thinking about expected value, the way I talk about it is relatively simple – in the sense of, say, you’re flipping a coin; the outcome is 50/50; you’re going to give me $2 for every $1 I risk; and so I’m making 50 cents on every flip, right?

That is because half the time I’m going to win $2 – so that’s a net gain of 50% of $2, which is $1 – and half the time, I’m going to lose $1, which is a net loss of 50 cents, because half the time I’m going to lose $1. And then I can subtract 50 cents from $1 and I get a 50 cent expected value on each dollar. Now, what’s important to note is there is no flip on which I will make 50 cents – in the sense that you are not going to put 50 cents in my hand. This is what I’m going to make over time, right? So if I bet it 10,000 times, I’m going to make $5,000 because I’ll make 50 cents on every dollar.

And when we’re thinking about expected value, we are thinking about interlinking problems, right? And this is something we think about in poker all the time ... I hope I’m understanding your question – I’m trying to answer the question I think you asked, let’s put it that way. So if I make a bet right now, it might be a break-even bet, if we just do that sort of coin-flipping math right now – but it’s actually a positive expected value, because it creates equity for me going forward.

This is actually a really important concept with bluffing, right? I could bluff you and the bluff itself – that moment – might not be positive expected value, if we only consider what the probability of me winning the current hand is, OK? But it is positive expected value because it is going to make it harder for you to play against me in the future. In other words, if you catch me bluffing, even though I lose the hand, I get something out of it, because it makes it harder for you to know – the next time I bet that way – if I have a really good hand or I don’t.

So you can see now, over time, with the decision I make right now, I have to include – is that improving my equity going forward or hurting it going forward? And that would then get included into the expected value equation. And you also can see that happening in parallel – like a decision I make could affect other decisions I have going on at the exact same time. In my books – because I’m not teaching people; they are not, like, risk management books – I am just trying to get people to start thinking about this concept of gains and losses. But obviously it is much more complex than that – because decisions are intertwined. So you have to sort of make your guess at how those things are going to be intertwined.

This is actually a really important negotiation concept, right? If I am negotiating and I never walk away, that’s really bad for me – not just in the sense that I’m probably taking bad terms, if I never break the deal; but also in the sense that it’s reducing my expected value going forward, because people then know I don’t walk away from negotiations. And so if I break a deal, it’s actually really good for me – because then, for the future, I get a reputational advantage for being a tough negotiator and that puts people in a position where they’re more likely to give me favourable terms, for example. So I hope that’s the answer to your question. That’s the one I understood you asking but I may have answered a completely different question!

‘Kill criteria’ and ‘pre-commitment contracts’

JTR: No, no – I think you absolutely did. Among the tools at our disposal to help us quit, you mention one in your book, which I personally think is quite powerful. Michael Mauboussin uses the term ‘man overboard moment’ and I think you have too – although in the book you use another name – ‘ kill list’.

AD: Yes – it’s ‘kill criteria’, which gets combined with a Ulysses contract and I call it a ‘pre-commitment contract’ in this particular book. ‘Kill criteria’ necessarily require a pre-commitment contract or a Ulysses contract so let’s go through that all step-by-step. A Ulysses contract is just a pre-commitment to a particular action and a really simple example of that would be, if I want to eat healthily, I can throw all the junk food out of my house. In other words, I’m not leaving the decision to when I’m actually in the moment – I’m thinking in advance about what my goals are, what the things are I’m trying to gain and then setting restrictions.

I’m binding myself to that decision in some way, which could be all the way from literally getting all the junk food out of my house, to telling somebody – I could say to you, Juan, I want to eat really healthily so, if you see me go for a donut, I want you to stop me. So that would be a pre-commitment contract. And it also could just be like an advanced plan – so I could just make a commitment to myself that, when I’m around junk food, I’m going to get an apple instead, right? So there are different things you can do.

Now, these pre commitment contracts – we know from the science – work really, really well. And why is that? Well, actually a really big problem with quitting is that as, Daniel Kahneman, the Nobel Laureate, phrases it: “The worst time to make a decision is when you’re in it.” And what he means by that is, when you actually have an open box of chocolates sitting in front of you, it’s really hard to make a good decision, because then there are all sorts of things that have to do with short-termism, say, and that’s when bias is going to be the greatest and we’re really in our own perspective – the things that we want to be true of the world. So it is harder for us to get outside of that decision and see it in some sort of rational way.

But when we’re thinking in advance, we’re actually much better at that – we’re much better at actually thinking pretty rationally about things. Now, how does this relate to the quitting problem? Barry Staw, who is just a giant in the field of this problem of ‘escalation of commitment’, said, you know, humans are supposed to have this intuition – that when we get signals from the world that we ought to walk away from something, then we will. You know, it’s that idea of, after we have started something, when we learn something new and it’s bad news, we will react to it and we will walk away. This was back in the 1970s and Staw pointed out this problem – he said, it’s just not true. In fact, not only do we not walk away, we escalate our commitment to the cause. In other words, we increase our commitment to the cause, which is what you’re seeing with these negotiation students – they keep going and changing their terms just trying to close the negotiation out, right?

So where pre-commitment contracts help us is we can pre-commit to certain actions – if we see those signals in the future. It makes a really big difference. What are the signals we’re looking for? Well, those are ‘kill criteria’ – so basically a list of the things you might see in the future, that would tell you it’s time to walk away. A simple implementation of that – which I’ve done with portfolio managers – is that portfolio managers have the intuition that, when you’ve done the quant work, you have a thesis you might be trading and that thesis is going to imply certain things about what you’re expecting of the world. And so they go on to intuit that when the fundamentals move against them in a certain way, for example, then obviously they’re going to get off the trade, because their thesis implied certain things. So, when it goes against their thesis, they will quit, right? But, in reality, the answer is no, they will not.

So what I’ve done with them is said, actually lay it out – what does your thesis imply? Let’s write it down and then let’s figure out what could be happening in the world that would tell you that you want to quit. Let me give you a super-simple example of kill criteria. So I think a lot of people bought crypto – let’s say, bitcoin – a lot of people bought bitcoin because their thesis was it would be a hedge against inflation. I’m not saying everybody bought it for this reason but there were a lot of people who bought it for that reason – that it was going to be a hedge against inflation.

And I’m not saying whether that was a good thesis or not – again, we make these decisions under conditions of uncertainty. And if you’ve never experienced bitcoin existing at a time when inflation is rising, you’re making guesses about the correlation between those two things. So I’m not going to quibble with the thesis but now the question is what happens when it is correlated with inflation – do people sell? And I think we saw the answer is no – I mean, it went down but, I think, out of necessity, right? But people come up with different rationalisations – well, I’m not really holding it for that reason or I’m holding it because I think it’s actually like a tech play or I don’t know, whatever. They come up with different things.

Now, let’s imagine you had kill criteria and a pre-commitment contract, right. The kill criteria would be something to the effect of, if I see inflation go up a certain amount, and I see some correlation occurring to some degree with inflation – and you would define what those values were as being enough to disprove your thesis – when that happens, I must sell. And it just turns out – and Barry Staw has shown this – it makes you much better at cutting your losses when you have thought about it in advance and before you’re already in the decision.

JTR: One example you gave in your book brings it back to Everest – the three people who survived one of the tragedies because they actually followed the criteria to turn back when they realised they were never going to make it to the peak in time.

AD: Yes. There are all sorts of really simple examples of ‘kill criteria’ – and one of them, by the way, is a stop-loss order. That’s a kill criteria. Stop-losses are something I used in poker – it is like, I’m going to be terrible at deciding why I’m losing when I’m losing! So let me just say, in advance, if I lose a certain amount, I’m just getting off this thing, right? Like, I’m getting off the ride now!

And another really simple example you point out is mountain climbing. When people are climbing mountains, they have something called ‘turnaround times’, which are essentially like stop-losses – like, if I’m not at my final destination by this time, it doesn’t matter where I am on the mountain, I have to turn around. So on Everest, on summit day, they have a turnaround time of 1pm: you leave Camp 4 to go up to the summit and the rule is, no matter where you are, whether you’ve made the summit or not, at 1pm you have to turn around.

Now we know people don’t follow these all the time. They are not perfect because people do die up on the mountain long past 1pm. But you’re more likely to follow them – and ‘more likely to’ is a lot better in this particular situation ... or in any particular situation. If you’re more likely to quit when you’re supposed to, you’re obviously going to have better outcomes over the course of your life. And there is a really great story of these three climbers, Hutchison, Taske and Kasischke, who were part of an expedition in 1996 of eight climbers, three Sherpas and an expedition leader. The trio had sort of become friends and were climbing together – they had sort of formed their own little group – and they set out for the summit. And in the 1990s, expeditions for Everest were really popular so there were a lot of people on the mountain.

Now, I also just want to set the stage and say they had paid something like $75,000 to do this expedition – so remember, when we were talking about sunk-cost problems, right? They have paid a lot of money to try to summit Everest – and then think about the time and the training. How much of their identity is tied into this, right? They’ve had to make these choices of, like, not spending time with their family and everybody knows they’re trying to climb Everest. So they have a lot of things stacked up against them that could make them keep going for the summit, even if the conditions weren’t good to do so. That is true for anybody who is climbing that mountain. Think about it, what’s worse – like, never thinking you want to climb Everest or coming within 300 feet of the summit and not making it? Obviously, the second is so psychically painful – even though you’ve climbed a lot – because, again, it’s this sunk-cost problem, right?

So, anyway, they are going up the mountain and it is really, really crowded – there are more than 30 people trying to summit in the same day and you have to summit a little bit in single file, so there’s a traffic jam. Their expedition leader comes up behind them and Hutchison stops him and asks, how long do you think it is until the summit? And the expedition leader says, well, I think it’s three hours. He then continues on and Hutchison holds Taske and Kasischke back and says, listen, we have a problem. We’ve been told the turnaround time for summit day is 1pm – and the reason is, if you summit after 1pm, you’re too likely to descend the mountain in darkness, which is super-dangerous. Hutchison tells the other two, it’s already 11.30am – so if it is three hours, we’re not going to get to the summit till 2.30pm.

So this is a very good example of forecasting, right? He is standing on the mountain with these other two guys, they have lots of oxygen and they aren’t exhausted – they are in fine fettle, right? They’re totally fine. But Hutchison says, I can tell we’re not going to get to the summit by 1pm so we should turn around now. There’s no reason to keep going at this point. They have a little confab, it takes a little convincing but they decide, yes, we’re going to turn around and we’re going to go back to Camp 4. And that’s the end of the story.

Now, there is a big reason why you might not know who these climbers are – this is part of the problem with grit, right? It is not a heroic story – at least, not as we traditionally think about heroism. But the interesting thing is, if people do know who these climbers are, it is because they were part of the expedition that was chronicled in Jon Krakauer’s book Into Thin Air; and in the documentary Everest; and in the movie called Everest. And they were part of Rob Hall’s expedition, who quite famously got up to the summit at 2pm. He is the one who told them it would be three hours, by the way – so he is a pretty fast climber. He did manage to get there by 2pm – but, notice, an hour past the turnaround time, which he knew because he had said it. And then Doug Hansen, his other client got there at 4pm, immediately collapsed and died. And then Hall could not get back down the mountain and he perished on the summit as well.

So, what about these three climbers? They are in the book and Krakauer says they were the best decision-makers on the mountain – but people tend not to remember them because we think about the people who persevere as the heroes of the story. But I want everybody to consider this for a second – in following the turnaround time and following the kill criteria, these three climbers had to turn around when everybody else was continuing on. They had to be willing to be OK with the fact they had spent $75,000 and they weren’t going to succeed. They had to face the possibility that everybody else got up to the summit just fine and came back down and they felt like idiots. They were facing all of that stuff – and yet they turned around anyway. So I would argue that was actually the more courageous choice.

JTR: I couldn’t agree more. I actually read the book many years ago and it is a mind-blowing story – and very tragic as well. Now, we had Joe Sweeney of the Alliance for Decision Education on the podcast at the start of the year to talk about the importance of teaching kids at a younger age all the tools to become better at decision-making. And I was thinking it must be difficult to find the right balance between teaching the importance of quitting and not quitting for the wrong reasons. So I wanted to ask you, what is the right way to teach children about this and at what age?

AD: I think you should start right away. A lot of this has to do with this idea of, how do you get out of the decision? Right? So, first of all, you definitely want to have kill criteria, exit criteria, quick criteria – whatever you want to call them – and you also want to have a ‘quitting coach’, which is someone to help you to see whether the thing you’re doing is still worthwhile or not. This is where I think, for young children, parents can become very helpful.

Where parents are unhelpful is they have a bias towards sticking – because we all do. Somebody told me a great story – his parents wanted him to play a musical instrument, they let him choose. And he chose the saxophone – I guess he thought it was cool. And when he started playing, it turned out he just really hated the taste of the reed in his mouth – there was something about it that was really unpleasant for him. His parents, because they wanted him to build character, did not allow him to quit and now he doesn’t play any musical instrument whatsoever – because it was such a traumatic experience for him to have to keep going.

So, in terms of being a quitting coach, you have to think about what the thing is you’re actually trying to help the child to accomplish – and communicate that well. So if I want you to play a musical instrument and you tell me the reed is really bothering you, the correct response should be: if it’s that bad, OK, but what instrument are you going to play instead? And, if they then come to you and argue they’re not musical and they’re tone-deaf – this actually happened to me – then it should be: OK, then what enrichment activity are you going to do instead? That way you make sure they’re accomplishing what you are trying to help them to achieve – as opposed to the saxophone being the goal itself.

So this happened to me. I played piano and I was tone-deaf, for real. I said to my mother, I’m terrible at this, I shouldn’t keep doing this and she wouldn’t allow me to stop. She made me do it for a year. And then, as soon as I was free, I switched to gymnastics, which was a much better activity for me. And I stuck to that till I was 14 years old – like, I was eight when this happened – because I was a gritty kid. It was stupid for me to be playing piano – and I think we lose sight of that aspect.

This is generally a problem with goals – we start out on something because we’re trying to achieve something broader and then we get stuck on the thing. Like, ‘it’s the saxophone’ – not, ‘it’s whatever you’re trying to achieve through the enrichment that playing a musical instrument might bring you’. And I think, if you can start thinking about that well, you can start to communicate that to your child well. And then, when your child is doing something you think is worthwhile and they enjoy, and they have a really bad day on the soccer pitch and they come off crying and saying they want to quit – that’s where kill criteria come in.

You can say to them, I think that was probably just a bad day – and everybody has bad days – but that doesn’t mean you should just quit right away. Let’s actually think about how many games you have to play in order for us to figure out if this is something you just really hate and it really isn’t for you. Let’s figure out how many games you have to give it – because there’s a learning curve and, at first, you’re going to be bad and then you’re going to get better. So let’s figure that out.

So maybe you say, let’s just play this season. Maybe you do that. And then you say, what are your feelings going to be? Do you think you’ll have scored a goal? How are you going to feel about being part of a team? You can figure out all the things you think would indicate this is a good thing for you to keep doing – and then what are the things that would tell you maybe it’s not the right thing for you to keep doing? So you can literally set those out with your child and then agree they’re going to stick to it long enough to be able to get that information. And then, once they get that information, we'll make a decision about whether they stick or quit. But part of that decision is going to be what are you switching to? Wouldn’t it be great if we all learned that lesson instead of being forced to play the saxophone when we hate it?

Could anything help Putin quit Ukraine?

JTR: Absolutely. This might be a bit of a cheeky question but when you need to save face, especially when the stakes are very high, and you have lots of eyes on you, does that make the decision to quit more difficult? In your book, you mention the case of American free-solo climber Alex Honnold, who in 2016 made the decision to quit on the day he had initially planned to climb the El Capitan peak [for the Free Solo documentary and which he ended up scaling the following year]. But let’s think about a decision where the stakes are even higher – with the Russian invasion of Ukraine, given nothing has gone as he clearly expected it to, what would help Vladimir Putin to quit?

AD: Yes, nothing – I mean, this is the problem. Interestingly enough, I mentioned Barry Staw, who is one of this generation of scientists who really started noticing this problem with escalation of commitment. They all appeared in the 70s – and that’s not accidental. There was something very particular that was formative for them, which was the Vietnam War, and this was a generation of scientists who saw America just mashed in this war – there was not one single point where anybody thought America was winning.

It was very clearly, from the get-go, just a completely losing proposition – and one where it was not just so much money lost. Think about how many lives were lost to that war – after the point they figured out they weren’t going to win it. There was something called the Pentagon Papers that came out at the time – essentially, internal Department of Defence papers – where it became very clear that George Ball, who was an undersecretary of state at the time, had said, don’t get started in this because you’re not going to be able to stop. Like, once you start a war, short of victory – whatever ‘victory’ is – you will not be able to get out of it.

So this was the inspiration for that set of scientists and we can see everything happening here, right? So there are sunk costs – you spent a lot of money. There is also the really heart-breaking sunk cost of the lives lost. You know, in regards to Afghanistan, General Tony Thomas said he would talk to Gold Star parents who would say, go win this war, because I don’t want my child to have died in vain. And as heart-breaking as that is – it’s obviously incredibly heart-breaking – but should the next child die? Do we have enough of a chance of actually achieving our objectives that the next child should die? But it’s very hard for us to think that way?

So that’s sort of the very classic sunk-cost problem but then you have this identity issue. Like, your national identity – how are people going to judge you? And the interesting thing is, taxpayers fall for the sunk-cost problem too. When people shut down public-works projects, they’re like, oh, what a waste of taxpayer money – but they may be shutting it down so they don’t waste $100bn more of your money. Yet we don’t naturally think this way.

You have this issue of ‘external validity’, as it’s called, which is how are others going to judge you? And, in this particular case, when you withdraw from a war without having achieved your objectives, you get judged very harshly, right? We saw that with the Vietnam War – it is part of the reason why people stay in a war and try to win it. Look at what happened with Afghanistan – one, two, three presidents before Biden had promised to get us out of that war, including, by the way, George W Bush, who started it, right? He promised to get out of it. But we didn’t. Obama promised to get us out of the war. But he did not. Trump promised to get us out of the war. He did not.

Biden promised to get out of the war. He did – and people are mad. They’re mad. And you could see what got revealed when we exited was we weren’t ever winning the war anyway, right? It took like three days for the Taliban to retake it, when it took us 20 years, and I don’t think there was a single person who thought we were achieving our objectives in Afghanistan. I think people felt we were achieving the objectives in Iraq. I’m not saying whether you agree with the objectives or not – I just think people felt like they were being achieved – but certainly not in Afghanistan at any point.

Now, let’s go to Vladimir Putin. The interesting thing with Putin is it is worse than America in Afghanistan – we had allies there with us so we weren’t standing against literally the whole world. Putin, at best, has some people who are tolerating what’s going on but he doesn’t actually have any allies here. The West is against him while the people who might be in his coalition are just sort of like, we’re not happy about it but I guess we won’t totally say that out loud.

His political identity is obviously embroiled in it. The national identity of Russia is. He had stated objectives of taking the whole thing – he was planning to take Kyiv in a few days. Obviously, there is not a single measure under which he is winning this war and where things are going well. And people keep talking about ‘off-ramps’ for him but, if you study the science, like, what off-ramp? He is supposed to abandon everything he’s put into it and just admit that, like, the West wins?

This is the problem – it is very hard for us to walk away. And again, when you think about, for example, withdrawing from the war in Afghanistan, there is no doubt you are going to get tons of flack for that – there’s no doubt people are going to be really mad. And so to do it anyway – to finally follow through on that promise – was actually pretty spectacular from the point of view of the biases against quitting. So there are just too many things lined up there that will make it pretty impossible for Putin to leave.

So what are the conditions under which Russia would leave? I mean, I suppose if they have no more people left to fight, that would do it – and they seem to be getting close to that. They have either conscripted everybody or they have all fled, right? If Vladimir Putin is deposed – because you can separate Russia’s identity from Putin here. So if you get rid of the person whose identity is tied to this war, Russia then can save face – because that’s what you need to create – and they can then get out, right?

So if there is some sort of internal regime change in Russia, they would be able to get out of the war. But as long as Vladimir Putin is in power, I just don’t see it from a scientific standpoint – any ability to be able to walk away from that. Unless, again, they are completely running out of resources – but I think, if they completely run out of resources, he gets deposed anyway so I think that becomes sort of recursive anyway. And that is the problem. I mean, when it comes to war in particular, one of the things that I have really learned from this book – from my own research – is, be careful what you start because, once you start it, it’s really, really, really hard to stop.

‘The hardest thing to walk away from is who you are’

JTR: In your book you discus how, if you create an association with an identity and are then very vocal about that identity, it is very difficult – potentially even impossible – for you to remove yourself from it.

AD: Yes – the hardest thing to walk away from is who you are. So how do things become part of our identity? Well, first of all, obviously, if they’re very public-facing. For me, I played poker on television so being a poker player was very much part of my identity – but that’s really true for any career we have, right? Like, ‘I’m a doctor’ – we don’t say, I practise medicine. I mean, some people do, but mostly you say ‘I’m a doctor’. So the things we do can become part of who we are – particularly when they’re public, which our careers are.

And particularly, here’s another one, when our beliefs make us stand out from the crowd. When we hold beliefs that make us stand out from the crowd, it becomes much more part of our identity – because one of the ways we just naturally define ourselves as humans is through both ‘belongingness’ and ‘distinctiveness’. So ‘belongingness’ is belonging to a group – that group identity. The group of poker players, for example, And ‘distinctiveness’ is, how am I different from other people? We can see this pretty clearly in politics, right? Like, you belong to one political party, which stands you in opposition to people – you’re distinct from the other political parties, right? So this becomes very much part of our identity under these circumstances.

So whenever identity gets involved, it becomes very hard to quit because we’re really asking you to then quit who you are – to say that is no longer what I am. Or even when it comes to a belief, to admit that was wrong is another way to have to realise a loss. And so what happens is, when the world gives us signals we ought to walk away from who we are, we’re very good at rationalising away those signals because it creates too much dissonance for us, right? It’s the world and us – like, I’m now in some sort of fight with the world and it becomes really too hard to resolve the dissonance.

So let’s think about how this manifests in a couple of ways. First of all, on the distinctiveness thing, I’ll give you a super-simple example. Let’s say you used to believe Pluto was a planet – the question is, does that really define your identity? Not really because everybody believed it – that’s the thing. When we all thought Pluto was a planet, were you arguing with anybody? No – everybody was saying Pluto was a planet. So when everybody is doing it, it’s not really identity-defining.

But what if you believe the Earth is flat? Now that’s identity-defining, right? Now you belong to a very specific group of people who believe the Earth is flat, as distinct from consensus. So you now stand out from the crowd. So now let’s think what happens when a bunch of scientists tell you Pluto isn’t a planet anymore. Everyone is like, OK, I guess I was wrong – because what are you giving up with that? Nothing, right? But when people tell you the Earth is not flat, it’s round, what do you have to give up to accept that? Now you have to give up a lot. I use flat-earthers as an example but people might think, well, they aren’t the best critical thinkers anyway – but, just to alarm everybody, it doesn’t matter! This is not a critical-thinking problem – it’s a ‘the way your brain is wired’ problem.

I mean, thinking the earth is flat is a critical-thinking problem but a reluctance to abandon the belief is different! John Beshears and Katy Milkman actually did a great study on stock analysts. OK, so now we’re talking about critical thinkers – these people are pretty rational, one would hope. And they basically took some 6,000 forecasts on earnings estimates they had put out, and they wanted to understand: what happens when they make an estimate, say, for the end of the next quarter, and the estimate just doesn’t bear out? So where the actual earnings are very, very different to the analyst’s forecast.

So they divided the 6,000 forecasts into two categories of ‘consensus’ and ‘non-consensus’ – so, in effect, one is ‘Pluto’ and the other is ‘Flat Earth’ – and then they looked at the actual earnings and asked, what happens when these conflict with the forecast? And it turned out that when they were ‘Pluto’ kind of forecasts, where it was totally in consensus and everybody agreed, and the actual earnings didn’t bear that forecast out, the analysts would change their mind. And when you looked at their updated forecast for the next quarter, they would change in the direction of the new information, like a rational person. So that is not that surprising.

But what happened when they made ‘Flat Earth’ kind of forecasts? Like, when they were way out of consensus and made a forecast that was really, really different that made them stand out from the crowd. Now, when the actual earnings didn’t bear that out, they doubled down on the forecast. They didn’t update it anymore. So that’s alarming! They had a financial incentive to be accurate so they were really going against their own financial incentive here in order to stick to their guns on what their belief was before – but, again, only when it makes it part of your identity.

Three book recommendations

JTR: You reference Katy Milkman in your book as well. She has a great podcast and she wrote a great book, How to Change. We have actually invited her twice onto the pod but she has been extremely busy – still, we hope she will come on at some point in the future. Annie, I am conscious of time. I have so many questions about your new book and how to think about quitting because it is a fascinating topic for us and so important for what we do. And I had never really thought about it as one of the most important elements of trying to become a good decision-maker. So thank you very much for that. Last time, when we asked for a book recommendation, you chose Maria Konnikova’s The Biggest Bluff – and actually, thanks to you, Maria was on the podcast last year. I was wondering if you could provide us with a new recommendation – it can be one of your own books, it can be something you have read recently or anything really.

AD: Well, I think people should read The Power of Regret by Dan Pink – selfishly, because I think it fits well with my book! You know, we think about regret as a really bad thing and, in investment, for example, financial advisers will do regret-minimisation strategies, which I personally think exacerbates loss aversion, which is a problem. And he’s really arguing that regret has a lot of good stuff associated with it – like, it’s how we learn, it’s how we figure out what our values are and what we want to do in the future. So we become better people for the things we regret, in a lot of ways. So that is a really good one people should read.

The other one that, again, just kind of philosophically fits with my book is Range: How Generalists Triumph in a Specialised World by David Epstein, which is wonderful. And then I have to recommend Grit by Angela Duckworth, if people haven’t read it yet – I mean, it is a huge bestseller. But I want to be really clear – I’m not telling people, you shouldn’t stick to things. I’m like a super-gritty human – of course, we should stick to things. It’s just that you have to understand both sides of the equation. I mean, chapter one of Quit is titled ‘The opposite of a great virtue is also a great virtue’. And I think that’s the best way to think about ‘grit’ and ‘quit’ – they stand together and they’re both really important for success. So I think I would be committing malpractice if I didn’t also recommend Grit – because I think you need to read both.

JTR: That’s fantastic. Annie, thank you very much for coming back to The Value Perspective podcast. Many congratulations on your latest book – it was fascinating and we totally recommend all our listeners read it – and congratulations for doing a full circle and finishing your PhD.

AD: Thank you. Thank you. Well, I haven’t finished yet but hopefully soon!