The Peter Attia Drive - #96 - David Epstein： How a range of experience leads to better performance in a highly specialized world

00:00:00.000 Hey everyone, welcome to the drive podcast. I'm your host, Peter Atiyah. This podcast,

00:00:15.480 my website and my weekly newsletter all focus on the goal of translating the science of longevity

00:00:19.800 into something accessible for everyone. Our goal is to provide the best content in health

00:00:24.600 and wellness full stop. And we've assembled a great team of analysts to make this happen.

00:00:28.880 If you enjoy this podcast, we've created a membership program that brings you far more

00:00:33.280 in-depth content. If you want to take your knowledge of this space to the next level at

00:00:37.320 the end of this episode, I'll explain what those benefits are. Or if you want to learn more now,

00:00:41.720 head over to peteratiyahmd.com forward slash subscribe. Now, without further delay, here's

00:00:48.080 today's episode. I guess this week is David Epstein. David is the author of the bestselling book,

00:00:54.160 The Sports Gene, and more recently, Range, Why Generalists Triumph in a Specialized World,

00:00:59.380 which ironically makes him a specialist of generalists. He was previously a science and

00:01:04.460 investigative reporter at ProPublica. And prior to that, a senior writer at Sports Illustrated.

00:01:08.920 I have followed David's work for some time, have enjoyed his articles, his books,

00:01:13.160 and reached out to him several months ago, asking to interview him, even though I had already heard

00:01:18.840 him on a couple of other podcasts that frankly, I thought did a great job, but I just knew there was

00:01:22.620 more I wanted to explore with him. So of course, if you've heard David on other podcasts, I wouldn't

00:01:26.220 let that discourage you from listening to him here. We go into, in my opinion, sort of a broader,

00:01:31.420 deeper discussion that's sort of enabled by our long format. We talk about a lot of things,

00:01:35.460 but I think at the root of it, it's basically trying to get a better understanding of how we

00:01:40.040 can improve our own performance. And perhaps you can see we get into a lot of stuff around kids,

00:01:44.880 but as parents, how do we manage the exposure of our kids to various things? I think anyone

00:01:51.520 listening to this who is a parent obviously feels pretty strongly about giving their kids the best

00:01:56.100 chance at finding something they love and doing well at it. There are so many things we go into

00:02:00.160 in this podcast that are just fascinating beyond belief, including a really good explanation of

00:02:04.220 why the 10,000 hour rule that most people take for granted is essentially an axiom or dogma,

00:02:10.380 i.e. that 10,000 hours of deliberate practice is what is required for mastery and greatness. And I think

00:02:15.060 David goes into a great explanation of why that's probably completely nonsense, which is not to say that

00:02:19.760 deliberate practice is not incredibly important, but to break it down to something as simple as 10,000

00:02:24.760 hours is almost assuredly incorrect. So without further delay, please enjoy my conversation with

00:02:30.600 David Epstein.

00:02:36.320 Thank you so much for coming.

00:02:38.140 It's a pleasure. Thanks for having me.

00:02:40.520 Your work has been something I followed for quite a long time now. And even before your most recent book

00:02:48.680 came out, just on the basis of the book you wrote, was it in 2013?

00:02:53.180 That's right.

00:02:53.880 Yeah. Just on the basis of that book, people had always said, oh, you got to interview him. And I was

00:02:57.660 always like, yeah, yeah, I know, I know, I know, I know. And then the new book comes out and it's

00:03:00.980 like, well, now there's no excuse not to. So, and I guess it's tough for you because you have this new

00:03:05.160 book and everybody wants to talk about it, but I kind of want to talk about the old one too, if that's okay.

00:03:08.180 Whatever you want. We can talk about something other than my books if you want, whatever you want.

00:03:12.140 We have a lot to talk about actually, because I learned a number of things about you in getting ready to sit down

00:03:16.520 today, but somehow it escaped me that you were a fan of IR10 Senna.

00:03:20.080 Yeah. I mean, first of all, one of the all-time great sports documentaries if people have seen

00:03:23.780 Senna, but I used to follow racing and just, I loved racing in general. Like if you put two

00:03:29.460 paper boats on a pond, you know, I was interested basically. And then as I learned about him and his

00:03:35.900 start in karting and the different types of racing he did, the fact that he was a good gymnast as a

00:03:40.580 young guy, that he basically tried to retire from racing at one point, you know, and that he was also

00:03:45.020 very quietly charitable. And I think, I think kind of a sensitive soul in a lot of ways, and obviously

00:03:50.660 a very dramatic story. And I think the greatest Def 1 driver ever. And so just a lot of, I don't know

00:03:54.900 nearly as much as you, so I'm kind of cautious about saying anything, but yeah.

00:04:00.000 I think all of those things are really interesting and resonant. And so last week I went to the beach

00:04:03.940 and it was with my kids and the two older kids were in the water. So just me and the toddler,

00:04:08.800 who's two, who's actually named after Senna, his name is Arathon. And then these other two boys

00:04:12.940 came up and started playing. And so then I'm sort of overseeing a play group, basically one of mine

00:04:18.500 and two others. And for like 20 minutes, this is going on. And then the mother of these two other

00:04:24.680 boys comes up and she's thanking me or whatever for playing with them. And I can sense a Brazilian 0.97

00:04:30.060 accent. I said, oh, are you Brazilian? She says, yeah. And I said, oh, great, blah, blah, blah, blah, 0.98

00:04:33.520 blah. And we get chatting. And I can't talk to somebody who's Brazilian for more than five minutes 1.00

00:04:36.920 without of course asking them, do you remember when Senna died and blah, blah, blah. And I'm still

00:04:41.840 waiting for the exception to this to happen. It's never happened yet where that person doesn't

00:04:46.520 immediately transform into, oh my God, Senna is the greatest manifestation and representation of 0.92

00:04:54.000 Brazil. It was the saddest day of my life when he died. And this woman would have been seven when he

00:04:59.900 died. And our nanny, also Brazilian, same thing, was like four when Senna died, but it's seared in her

00:05:06.100 mind. So I'm talking Uber driver, person at rest, it doesn't matter. Anytime I meet someone from Brazil, 1.00

00:05:11.720 we talk about Senna and it is without exception, they speak of him with a reverence that I don't

00:05:17.840 think Americans can relate to. There is no athlete we talk about. There's no politician. There's no

00:05:23.300 scientist we can speak about in the way a Brazilian talks about Senna.

00:05:27.480 That's amazing. So it sounds like what people here of certain generation might think of where

00:05:31.560 they were when Kennedy or Martin Luther King were assassinated or something like that.

00:05:35.060 It's absolutely that way because of course, everybody in Brazil was watching F1. So the world

00:05:40.780 stopped every Sunday to watch the race. So now you take, everybody is there stopping to watch

00:05:46.300 your guy doing this thing. And then you see this person die. But unlike maybe with say JFK,

00:05:53.220 where yes, anybody who was old enough at that moment would remember it. I don't even think that

00:05:57.780 they can speak about JFK the way, with as much love or reverence as they do Senna. So it's kind of

00:06:03.200 amazing to me and humbling.

00:06:04.540 That's interesting because obviously we have Ali as the first person who would come to mind.

00:06:07.600 But in his day, a lot of people hated him. It's a lot of athletes like him. I think

00:06:13.100 they become beloved once they're sort of older and non-threatening feeling. And so a lot of people

00:06:19.460 who love them don't actually really know what it was like for them at the time. But that's

00:06:23.540 fascinating. And I think his reputation, I think a lot of the, it turned out after Senna died that

00:06:28.880 he had been doing like a lot of charity that people didn't even know about with no,

00:06:33.040 no fanfare whatsoever. And so I think his legacy was, and that's amazing. That's not very,

00:06:38.340 very common.

00:06:38.920 That's right. Very few people know about how much he did for, for the people in Brazil and how

00:06:44.400 seriously he took his position of, he came from privilege, came from a wealthy family,

00:06:50.160 achieved this unbelievable success. And people love this humility that he had that,

00:06:55.420 because remember, it's not like he was the first Brazilian F1 champion. I mean,

00:06:58.380 Fittipaldi was a two-time champion, Nelson Piquet, a three-time champion, who, by the way,

00:07:03.360 Senec won his first championship the year after Piquet won his last. So it's like a complete overlap

00:07:09.780 with another, but they don't even belong in the same sentence for most Brazilians. In fact,

00:07:14.340 I asked this question of almost everybody as well, which is, how does he compare to Pelé?

00:07:19.880 And they're like, oh, Pelé was great, but.

00:07:22.240 Oh, really?

00:07:22.920 Oh, absolutely.

00:07:23.520 Oh, that's interesting.

00:07:23.780 Yeah, yeah. It's really kind of amazing.

00:07:26.680 Yeah, I guess that's the most telling question you could ask in Brazil, probably, right?

00:07:29.520 Yeah. I'll be going for my first time, actually, to Sao Paulo soon to watch an F1 race and to go,

00:07:34.920 and I want to be able to visit the memorial where he was buried and go to the foundation and stuff

00:07:39.440 like that.

00:07:40.040 And how did you get so interested in him?

00:07:41.920 I don't do anything in moderation, I think is what it kind of comes down to. And I've always loved

00:07:47.200 and been attracted to people that are incredibly passionate and great at what they do. And I do think

00:07:52.880 that his perfectionism, well, I'll take a step back and say, I was probably attracted to things

00:07:59.500 about him that I didn't appreciate the pathology in at the time. So I do think that his desire to

00:08:06.180 win probably also killed him. And I think that the sharpness of that edge, I probably found incredibly

00:08:13.660 appealing in a way that almost maybe speaks to my own demons. And I think that's probably true of a

00:08:19.660 lot of people. I don't think I'm unique in that. So I just remember, one of the things I remember

00:08:23.860 loving about him during his career was how much he cared about the engine and what was going on

00:08:29.400 with the car and the setup and the time he would spend with the mechanics. I mean, it was always

00:08:34.100 telling to me that the Honda mechanics loved him. I mean, just loved him. You'd have some guys that

00:08:40.120 would show up, they would drive and they would leave, but not Senna. Like he could spend the entire

00:08:44.260 night in the garage, machinating over every minute detail of the car. So it was just this sort of

00:08:50.640 incredible degree of perfectionism. Also, I do think that there was just a certain,

00:08:55.860 there were just things that he did that to this day can't be explained. I think his qualifying

00:09:00.700 session in 1988 at Monaco, there is no explanation for what he did that day. I'm sure you're familiar

00:09:06.040 with it just for the listener. Monaco is a very short circuit. So in a short circuit, the difference

00:09:11.280 between qualifying times should be tenths of seconds, hundredths of seconds. His teammate

00:09:16.700 that year, meaning someone driving the exact same car, which they had the best car in the field,

00:09:20.920 was Alain Prost, who was himself a three-time world champion at the time. Unbelievable driver.

00:09:27.080 Some would argue one of the more underrated drivers ever. Senna out-qualified him by a second

00:09:33.680 and a half. A second and a half on Monaco in a quali might as well be a day.

00:09:39.760 All right. It's like winning the hundred meters of the Olympics by a second and a half.

00:09:43.080 That's exactly right. It's like even Usain Bolt at his most dominant couldn't win a race by a second,

00:09:48.100 even though there's no actual onboard footage of Senna during that quali lap because he was already

00:09:53.420 on pole. So I don't think the networks were even paying attention to his very last quali lap, which

00:09:58.440 why would he try to go any faster? He'd already secured pole. But when you watch Senna at Monaco over

00:10:04.660 and over again, which is one of the most demanding circuits because of how tight it is,

00:10:07.780 I have my kids watch these videos because I'm like, I don't think you guys understand.

00:10:12.460 You think daddy drives a race car and that's fun because he can go fast. But I want you to see

00:10:16.480 what the best in the world is seeing in real time because we can't do this. Humans can't do what

00:10:23.600 he's doing. The other sort of extension to that story that speaks to this sort of love I have is

00:10:28.840 the tormented nature of this, which is what most people don't realize is he qualifies first in

00:10:33.480 Monaco or what most people I think have forgotten is he qualifies first from Monaco in 1988 by

00:10:37.780 literally a second and a half as the race is going on and on and on. He has built up such a lead. He

00:10:43.960 almost has a lap lead over the field with a very short duration to go in the race. I don't remember

00:10:49.180 how many laps, I think like maybe six to 10 laps to go. He could basically stop, get out of his car,

00:10:56.880 get back in it and still win the race. But he's pushing very hard. He's pushing so hard that he

00:11:02.620 actually crashed. He is disgusted with himself. He gets out of the car, literally leaves, goes

00:11:10.260 straight to his home in Monaco, doesn't speak with anybody for days. And to me, this is a guy for whom

00:11:16.560 it's not about winning. Yeah. Yeah. This actually gets to something. You said I could be digressive.

00:11:21.240 So I'm gonna make a multi, multi jump. Let's do it. Something I've been thinking about that I used to

00:11:25.280 think about a lot and then came up recently was at a certain point when I was at Sports Illustrated

00:11:30.060 and I was doing reporting on doping and I would get a lot of reader feedback of why are you reporting

00:11:35.100 on this? You're sort of a killjoy, that kind of stuff. I took it seriously and started thinking

00:11:38.920 about should I be doing this? What's the value that comes out of sport? And somehow I landed on

00:11:44.240 this book called The Grasshopper by a Canadian philosopher named Bernard Suits. And it's called The

00:11:50.080 Grasshopper because it's sort of a inversion of this Aesop fable where there's a grasshopper who's

00:11:55.000 playing games all summer. And while the ant is storing up food in the summer and then come

00:11:59.920 winter, the grasshopper doesn't have any food and the ant does. And the grasshopper goes to the ant

00:12:03.320 and asks for some food. And the ant says, no, you were playing while you should have been collecting

00:12:06.240 food. And so moral is kind of obvious. But in Suits, there had been this philosophical debate

00:12:11.240 that was supposed to be settled by Wittgenstein about is there any necessary and sufficient core

00:12:18.060 of sports and games? And he said, no. He said, no, absolutely. There's not.

00:12:21.280 And Suits, in writing this book, The Grasshopper, the grasshopper is a character who's playing these

00:12:26.080 games. And his disciples are come saying, you should be storing food. You're going to die.

00:12:30.040 And he says, no, this is who I am. I understand what's coming. But this is the best thing I can

00:12:34.920 be doing, this endeavor, for the love of what he was doing. And Suits says there is a core to all

00:12:40.440 sports and games. And it's the voluntary acceptance of unnecessary obstacles, which I thought was kind of

00:12:45.680 amazing. And he talks about what he calls the lucery attitude, which is the attitude you adopt

00:12:51.100 when you get involved in these things, which I think is kind of a love of difficulty, essentially.

00:12:57.580 And I think he sort of united something. Aristotle had these two, he put actions into two categories.

00:13:02.720 One was kinesis, which is like, build a house. You're doing it for the end. And the other was

00:13:07.200 energia, which is something like philosophical contemplation. You're doing it for the doing,

00:13:12.780 not for the end. And he said, these two things have to be separate. And I think what one of the

00:13:17.000 things Suits was saying was in sports and games, these things are united. There is an end that

00:13:20.800 you're going for, but the love of difficulty in the middle is what's really important. And you're

00:13:25.700 always doing something, you're intentionally doing something inefficient, right? Like you could walk a

00:13:30.720 ball and put it into a goal. That would be the most efficient way to do it. Or you could cut across

00:13:35.220 the track and get there faster, but you're intentionally engineering an inefficiency in order to

00:13:40.020 facilitate a certain experience. I'm sure Senna had that love of difficulty. And one of the things,

00:13:44.800 not to tie it to my own stuff, but one of the reasons I think about it is, to me, I don't know

00:13:49.500 that readers would say this, but to me, one of the major themes of my new book is that sometimes the

00:13:55.800 things you can do to cause the most rapid apparent short-term progress can undermine long-term

00:14:00.360 development. And that actually you don't always want to be as efficient as possible. And I think

00:14:05.040 that's very much embodied in this love of difficulty in sports and games, where you are

00:14:09.380 intentionally engineering in inefficiency in order to facilitate an experience that you hope has some

00:14:14.780 value and some learning. So sorry, that was my multi-jump. No, that is so true. I think that

00:14:18.920 actually in one sort of story captures the essence of the greatness we see in sports, which is, as you

00:14:26.100 said, it has to have a struggle in it. It's not interesting if there's no struggle, but it's in

00:14:30.500 service of some destination that can be quite arbitrary, by the way. I mean, race car driving happens to be

00:14:35.200 one of the less arbitrary ones. Going fast seems somewhat understandable and innate. Mountaineering

00:14:40.260 seems somewhat understandable and innate. Get to the top. But many sports, like basketball and

00:14:44.280 football, are kind of arbitrary in what we're asking people to do. Think of baseball. It's like

00:14:47.920 bananas. It's like if you were just watching it with no sense of the conceptual structure,

00:14:52.300 it would look ridiculous. Like, why doesn't the guy just stay at home plate? They're already there.

00:14:56.600 Yeah, that's so funny. Well, before we get to your new book, because there are so many questions I

00:15:04.320 have on that theme, both personally and then with respect to my profession, and then even more

00:15:10.040 broadly, I do want to go back to the gene. Because I remember when the book came out, I think it came

00:15:14.940 out on the heels of an article you had written in Sports Illustrated, correct?

00:15:18.020 Wow, I'm surprised you remember that. Yeah, the sports gene. Yeah, it did. Yeah. I want to confess

00:15:21.700 something really quick. This that nobody really calls me on is I wrote that article in Sports

00:15:26.220 Illustrated. It passed fact checking at Sports Illustrated because when the fact checkers called

00:15:31.460 back the scientists, they said, oh, this is true, this is true. But then after a year of, before my

00:15:36.080 books the first year, I try to just read 10 journal articles a day every day for the first year, no

00:15:39.360 writing. And having done that for the year researching the sports gene, I realized that while I had quoted

00:15:45.580 these scientists appropriately, some of them had told me things that could not be concluded from their

00:15:49.940 data. And so I cited my own article as one that was mistaken, but nobody really called me on that.

00:15:54.300 But that's, I think if you're writing about science, something you're writing about is going to be

00:15:57.800 wrong. So you have to kind of be ready for that. Well, it's so funny you bring this up. So Bob Kaplan,

00:16:02.140 who's my head of research, we are in the process now of going through the fact checking for this book

00:16:06.800 that I've been painfully and slowly working on for more time than I care to admit. And what we realized

00:16:12.840 is he can't be the fact checker, nor can I, because it's not just facts we're checking,

00:16:17.620 it's interpretation. And we are already so biased by our view on this. So we actually have another

00:16:23.580 one of our analysts doing the fact checking, but we specifically refer to it as fact plus

00:16:28.680 interpretation, fact plus interpretation, which turns out to be really a long process and a very

00:16:33.760 challenging process because you do, I mean, I've done this a handful of times where you pull up one

00:16:40.280 of the citation classics in medicine or science that people have referenced so many times. It's been

00:16:46.300 triply referenced internally to the point where I don't think the people referencing it anymore,

00:16:51.260 even know what the paper says, let alone what it's citing. Telephone game of citations. It's

00:16:55.640 unbelievable. And the few times I've had our team extract from those papers, I've been mortified at

00:17:03.680 how wrong they are, which again, they're not necessarily orthogonally wrong, but they've missed

00:17:09.120 so many things. Like it's like, oh, well, of course this so-and-so does such and such and such and

00:17:13.580 such. Well, let's go back and look, wait a minute. You realize that was in one really,

00:17:16.600 really, really bad experiment in mice in which you could never make that inference into another mammal.

00:17:23.460 And now yet it's taken as sort of a fact. I have to say not that that's good, but I have noticed

00:17:28.880 that that sort of thing provides opportunities for people like me where I'll go and read the

00:17:34.340 original research of things that have just been at the core of other bestselling books. It's kind of a

00:17:38.780 great, if you're willing to do it, it's sort of a competitive advantage. You know, it gives an

00:17:41.760 opportunity, not that I want people to be citing things wrongly, but I think you're totally right.

00:17:46.640 I started as a fact checker at Sports Illustrated and that's where you realize how many ways there

00:17:50.140 are to go wrong. Are you looking for any more work right now? No, no, no. I was happy to get out of

00:17:54.780 my fact checking days. I'm in my post book, never again phase, which I was in before, but that's where

00:18:00.380 I am right now. No, but yeah, I hired independent fact checkers also. And that doesn't mean there

00:18:04.560 aren't things that are wrong or interpretations that are wrong, but it certainly, certainly cuts it down

00:18:09.300 compared to, I think most books probably have no fact checking at all. Yeah. I'm super paranoid

00:18:14.060 about it because I also realized that we can't catch them all. That's the difference between a 0.71

00:18:17.480 blog post and a book is, you know, I've written more blog posts than I'll ever be able to count.

00:18:21.660 And the good news is the week it comes out, someone smarter than you is going to catch something that

00:18:26.720 you did wrong. And you're like, oh my God, yeah, totally right. Thank you for that. Boom. I can change

00:18:30.580 it. I can't do that with a book. And that is crippling me. Yeah. I mean, you can do small stuff for

00:18:35.800 second printings and things like that, but it's not as easy, right? It doesn't happen right away.

00:18:40.520 And if it, well, especially if it's the interpretation, it's one thing, if you get a

00:18:43.140 fact wrong, when you start to interpret something incorrectly and you come around, it's very difficult

00:18:49.080 to unwind that. Definitely. And if you're going to change, yeah, as I learned this time, the,

00:18:53.740 both of my books, there are 352 pages, I guess, if you count the front and back and because they

00:19:00.800 get printed in sets of 16. So everything has to fit to a multiple of 16, including the index and

00:19:06.920 the citations and everything. So if you have to change something major, like an interpretation,

00:19:10.520 where it's not one sentence, if you're going to mess up the page flow, it's not so doable.

00:19:15.480 You're making some unhappy folks. You alluded to something there, which is

00:19:18.840 the opportunity to go back and look at something that people have sort of taken as dogma and

00:19:23.040 questioning it. And in many ways, that's a big part of what the gene does and what your current book

00:19:27.240 does. I'll share with you sort of my bias coming into this discussion, not this discussion with

00:19:32.160 us per se, but, but sort of this theme, which is, so Daniel Coyle wrote a book in 2009 called

00:19:37.840 The Talent Code. And before that, there were a number of other books and pieces of literature

00:19:42.220 on that subject matter. And I was obsessed with this. I was obsessed with this idea of how can one

00:19:47.300 be great? Obsessed with this even as a child. And certainly when I was doing my surgical training,

00:19:52.760 I really remember spending lots of time reading literature on technically achieving mastery.

00:19:59.840 Like what does it mean to be a great musician or a great surgeon or a great athlete? Things where

00:20:05.420 there's some sort of dexterity and skill required that goes beyond just thinking and cognitive prowess.

00:20:12.100 And so I would say I completely bought this idea that deliberate practice is the only thing that

00:20:18.680 matters. And I think a lot of people sort of have taken that to be the case. What made you question

00:20:25.020 that in the first place? Or did you not question that and instead stumbled organically into questioning

00:20:30.320 that?

00:20:30.980 I did not question it in the first place. And I should say Dan's a friend and I'm a big fan of

00:20:34.340 his writing also. And yeah, I did not question it. In fact, if you saw my book proposal for The

00:20:40.720 Sports Gene, The Talent Code, in the book proposal, sometimes you do a section that, I don't know if you had to do

00:20:45.960 this or not, but other books that yours will be like. Other books like, yes, yes, yes.

00:20:49.320 I skipped that part for my second proposal. But in the first one, I had to do it. And Talent Code

00:20:53.080 was one of the ones that I said it was going to be like. And obviously, I would say for the casual

00:20:57.440 reader, it looks like they're actually diametrically opposed in many ways. And I don't see them as

00:21:02.140 diametrically opposed, but there are certainly some differences. And it was probably when I went back

00:21:08.460 and started looking at some of the original literature. The 10,000 hours rule was, who was I to question

00:21:13.340 that? I mean, the one good thing is I was in my past training to be a scientist. And I was like

00:21:17.380 living in a tent in the Arctic when I decided for sure to become a writer. So I knew I should leverage

00:21:21.200 that background of, I was in the geological sciences, which are pretty methodologically

00:21:25.060 rigorous, I would say, as the sciences go. And so I decided to go look at these original papers if I'm

00:21:29.760 going to study them. And I come across the first, the so-called 10,000 hour study that the scientists

00:21:34.860 who wrote it wouldn't call it that. But this was with violinists? Violinists, yeah. 30 violinists,

00:21:39.140 Famous Music Academy in Berlin, split into three groups. The top 10 who were deemed to potentially

00:21:44.480 be international soloists practiced in deliberate practice, highly focused error correction focused

00:21:49.700 practice, on average 10,000 hours by the age of 20. The first thing I noticed was that there were

00:21:57.000 no measures of variance reported in the study, which is not something when I was a grad student

00:22:02.260 that one could have gotten away with, reporting no measures of variance. So I was-

00:22:05.580 Explain that some folks might not even know what that means. So like, there's an example.

00:22:08.780 So if you look at a table and it says, this person practiced this many hours, this many

00:22:12.980 hours, this many hours, what was missing in that description?

00:22:15.220 First of all, no range. So several of the books and the paper wrote that there was complete

00:22:19.720 correspondence between the number of hours of practice and what group someone fell into.

00:22:24.140 And I said, well, I can't tell that from this data. Like, maybe someone in the lowest group

00:22:28.260 actually practiced more than someone in the highest group, but you haven't included the range

00:22:32.400 of practice hours or the standard deviations of practice hours. What is the individual variation?

00:22:36.880 Any time you take an average, it could be that nobody practiced 10,000 hours. It could be that

00:22:41.820 somebody practiced 100,000 hours and a bunch of people were much less.

00:22:45.660 So sort of like, what's my average, your average, and Bill Gates' average wealth?

00:22:50.180 Sure, right.

00:22:50.900 Sometimes averages can be wildly misleading.

00:22:53.380 That's right. I mean, so for example, in the chess literature, it takes 11,053 hours on average

00:22:57.920 to reach international master status. So 10,000 hours would be low. That's one level down from

00:23:02.120 grandmaster. But some people have made it in 3,000 hours. And some people finished a study

00:23:06.260 at 25,000. They still hadn't made it. So we don't really know where their endpoint is.

00:23:10.500 So you can tell someone, well, it takes 11,053 hours on average to reach international master

00:23:14.980 status, but it doesn't tell you anything about the breadth of actual skill acquisition.

00:23:19.400 So how is that possible, by the way? I mean, I can't imagine looking at a paper that wouldn't

00:23:22.820 at a minimum include a standard deviation for that type of calculation.

00:23:26.220 Don't know. Okay. So eventually I organized, I noticed that the most famous researcher on

00:23:31.560 that paper, who I think has done some very interesting work, and especially in the area

00:23:34.440 of memory, some work that I myself have tried to incorporate into things I do. But I noticed

00:23:38.620 that he was, in a lot of his work, saying there's no such thing as talent, it doesn't matter,

00:23:43.140 just pick any random thing and you'll be great at it.

00:23:45.280 Provided you put the work in.

00:23:46.260 Right, right, exactly. That it doesn't matter what you match with. And I noticed he was citing

00:23:50.560 a lot of physiology papers that I knew something about, like sports physiology papers,

00:23:54.300 and not, like you said, not in the way, the interpretation, it was kind of like these

00:24:00.200 secondary interpretations, telephone game stuff. And so I organized a panel at the American

00:24:04.520 College of Sports Medicine and invited him. And this was, I thought, a problem. He was

00:24:08.900 citing a lot of their papers. His work, Anders Ericsson, was super influential in expertise.

00:24:14.800 I think that 10,000 Hours paper is clearly the most influential paper ever in the development

00:24:18.420 of expertise. But they weren't talking to each other. So I organized this panel. And in that,

00:24:22.500 a researcher stood up named Tim Lightfoot and asked, what's the variance around that 10,000

00:24:27.180 hours? And he said, well, that doesn't really matter because the people were actually inconsistent

00:24:31.480 on multiple retrospective recalls. Because what they did is they just asked for retrospective

00:24:35.400 recall and then had the performers keep a diary for a week and then extrapolated it, basically.

00:24:40.060 Which, by the way, don't even get me started on the noise that's introduced by both of those

00:24:44.540 decisions.

00:24:45.140 Right. There was just a replication attempt, by the way, last month it published and failed. But we can

00:24:49.260 talk about that if you want. But in that, there was actually, in the new replication attempt,

00:24:52.780 there was someone at 4,000 Hours who got to the highest group and someone at 11,000 Hours who was

00:24:57.060 still in the lowest group. But anyway, so he asked, what's the variance around that 10,000 Hours?

00:25:02.020 And Ericsson says, he said, first of all, there was inconsistent recall. And so Tim says, yeah,

00:25:08.260 a lot of us struggle with imperfect data, but we still put measures of variance. And so Anders says,

00:25:13.660 well, that'll be like more valid when we have video diaries and we can really track it because

00:25:19.000 we're not being that precise anyway. And he says, again, we all struggle with imperfect data,

00:25:24.000 but we include measures of variance. Was it? And then he asked, so what was it? And he says,

00:25:27.480 I don't know. I'd have to go look back. And so Lightfoot says, definitely more than 500. And

00:25:32.520 that's where we leave it. And then I think two years ago, something, a couple of years ago,

00:25:36.980 Ericsson did publish measures of variance. And it turns out there was enormous variance.

00:25:39.800 Not only was there enormous variance in the original paper, the papers from 1993,

00:25:44.660 and it was three or four years ago that he finally published the variance,

00:25:47.920 made clear that their conclusions were wrong, that there was not complete correspondence between the

00:25:52.780 number of practice hours and the group that someone was in. Well, I think that's sort of where I'm

00:25:56.220 going with this question, which is you can't even make an observation of statistical significance

00:26:02.320 without variance. So I don't really understand what the paper is saying. This is the first thing

00:26:06.980 where I was reading it and saying, something's not right here. And so then I started asking these

00:26:11.020 very basic questions. Because I'd been an athlete, I'd gone from being like the worst walk-on on my

00:26:15.240 college team to being like a university record holder. So I'm like, yeah, maybe if I had trained

00:26:19.560 even more, I would have been even better. And that probably is true. But I was, I was like, oh, okay,

00:26:24.480 so there is no such thing as talent. I was convinced for a while. And then when I started seeing this,

00:26:28.500 I started asking the very basic questions like, okay, in my third year of training,

00:26:33.540 I could break the women's world record. So there has to be at least some basic genetic 0.99

00:26:38.720 difference because I haven't worked harder than the women who are pros by any stretch of the 0.99

00:26:42.440 imagination. And so I said, okay, let's start with that basic question. I'd contact sort of some of

00:26:46.440 the, I shouldn't call them the 10,000 hours researchers because they kind of disavowed that,

00:26:49.960 but the deliberate practice framework. And I remember contacting one and saying, wouldn't you agree

00:26:56.180 that a man and a woman who practice the same, like the man has advantages, which is why we separate 0.99

00:27:01.380 sexes in sports. And she sort of hedged and said, maybe not if they all train the same. I said,

00:27:08.280 really? And so she sent me a paper saying, in fact, we think this applies to other organisms.

00:27:12.860 If you look at this paper about racing dogs, you'll see that they practice the best ones in

00:27:18.060 the highest class practice about the equivalent to their lifespan of 10,000 hours. And so I'm reading

00:27:23.860 this and I started reading all the citations. And one of them notes that like half of these dogs

00:27:28.000 have what is otherwise an incredibly rare myostatin mutation.

00:27:31.820 Let me pause for a moment and explain to the listener what myostatin mutations are. So if you

00:27:35.200 knock out the myostatin gene, you look like a bodybuilder. Myostatin is a gene that inhibits

00:27:39.980 muscle growth. And there are lots of myostatin mutants out there that all have hypo-functioning

00:27:45.020 myostatin and therefore are super muscular.

00:27:46.960 Yeah. And so racing breeders had been, they didn't know about the gene, but they were breeding for fast.

00:27:52.320 They were clearly selecting for this trait.

00:27:53.860 Yeah. Yeah. And what they wanted was a single myostatin. If you get two, then you have a bully whip it and

00:27:57.720 Google that. It's pretty cool to see. Bully whip it.

00:27:59.660 That thing probably can't even move. It's so big.

00:28:01.440 Right. So they want the single mutation.

00:28:03.060 Yeah. Yeah.

00:28:03.340 And so I'm like, okay, most of these dogs...

00:28:05.260 This doesn't even mean anything.

00:28:06.220 So I wasn't going to like use this study or anything, but I just started saying,

00:28:09.160 these people aren't reading the primary stuff that they're citing or they're not tracking the

00:28:12.760 references back. And so I started to have doubts.

00:28:14.720 Is this a broader problem with non-experimental science?

00:28:19.280 I think so. 0.86

00:28:20.000 Because we do the same dumb thing in medicine, by the way. People sort of think of medicine, 0.97

00:28:23.280 which we're going to get to in spades as it's so rigorous. And yeah,

00:28:27.320 sometimes medicine does get to leverage the scientific method and actually get to do what

00:28:31.940 Francis Bacon talked about. But a lot of times you don't. I mean, when you think of some of the

00:28:36.420 most important public health measures that are out there, oftentimes they are based on exactly the

00:28:42.780 type of inference you're being appropriately critical of, which is observational, heavily selected,

00:28:49.840 range restricted, which I can't wait. You talk about this so eloquently that I cannot wait

00:28:54.600 to have you go off on your tangent soapbox, whatever, rant on that problem. But these are

00:29:00.160 huge issues.

00:29:01.780 Yeah. Yeah. I mean, that was the other thing. I don't know if I should skip to range restriction.

00:29:05.880 Well, let's start with medicine. So I wrote this one article when I was at ProPublica called

00:29:08.980 When Evidence Says No and Doctors Say Yes. And I should say, I love the medical profession. I think

00:29:13.780 it's filled with a variety of people like any profession, but a lot of people who really care and got into the

00:29:19.660 profession because they want to do something that is challenging and useful. But also there's like a

00:29:24.340 lot of poor science and there are a lot of things that continue to be done even once evidence shows

00:29:27.940 they don't work anymore. What Mike Joyner at the Mayo Clinic always calls bioplausible, something that

00:29:32.760 clearly makes sense. It definitely should work. It's just that when somebody does like a randomized

00:29:36.460 controlled trial, it doesn't. You probably saw the finished study of partial arthroscopic meniscus

00:29:42.920 repair. I think there's a lot of devils in the details. That said, it was interesting. They gave some

00:29:47.360 people sham surgery where they basically made an incision, banged around like they performed

00:29:51.060 surgery and sent them to physical therapy. And they did as well as the people who were getting

00:29:54.660 the surgery, which is mind boggling because everyone's doing something that seems like it

00:29:58.200 has to work. Someone's got knee pain. You bring them in, you give them imaging. They've got a tear,

00:30:03.180 fix it. How could that not work? But then I guess it turns out that some huge number of

00:30:06.640 people have incidental tear that doesn't have anything to do with the knee pain.

00:30:11.060 The meniscal tear is a huge thorn in the side of the orthopedic specialty because frankly,

00:30:16.600 I don't know the answer. I mean, my intuition is that that's a procedure that is probably done

00:30:20.320 far too much, but it's also probably a procedure that if you knew how to select the right patients,

00:30:24.560 you could probably make a difference. But because we don't, we end up applying the tool

00:30:29.060 far too broadly and we dilute the outcome. I'll give you a much more specific example that is so nerdy,

00:30:34.560 but there's a drug called ezetimibe or Zetia, which blocks cholesterol reabsorption. So the body

00:30:39.620 makes a ton of cholesterol. Virtually all the cholesterol in the body is made by the body and it gets

00:30:43.720 recirculated throughout the body. Well, part of this recirculation pathway requires that cholesterol

00:30:47.840 be dumped into along with bile into the gut. And then in your gut, you can reabsorb it. And the

00:30:53.920 body has a way to regulate how much of that's happening. But it turns out there's a drug that

00:30:58.380 blocks this thing called the Neiman Pixie one, like one transporter that drags cholesterol back in.

00:31:03.620 Now, when that drug is given in monotherapy, it lowers cholesterol, but not that much. And it

00:31:10.240 doesn't save lives. So it's not a drug that's really, in fact, it's absolutely therefore not

00:31:16.340 considered a first-line agent and it's never considered something that should be used in

00:31:19.620 isolation. Now, when you give it with a statin, it turns out it lowers cholesterol and it reduces

00:31:23.800 events. So the things that you care about, the actual hard outcomes change.

00:31:28.360 And it's not just the statin.

00:31:29.820 Correct. That's right. Because you can compare it to statin versus statin alone. So I have probably

00:31:36.020 kind of a contrarian view on this, which is I actually think this drug alone would work if you

00:31:41.420 actually only gave it to people who were hyper absorbers. But that's never been done. Because we can

00:31:46.300 measure how much absorption capacity a person has, but that's a kind of advanced measurement. You

00:31:51.580 wouldn't normally do that in a clinical setting. But if you select for patients who have mega

00:31:57.120 amounts of absorption, it's certainly possible that those patients... So I don't know the answer

00:32:02.180 to this. And only if a trial was done testing that way could you get it. But I do think that

00:32:06.980 this problem exists in medicine, which is you dilute by taking such a heterogeneous population

00:32:12.700 to test an intervention on. And you're therefore not really powered to detect an effect because

00:32:19.700 in your power calculation, you're using the entire population as your denominator. And really,

00:32:25.160 it probably needs to be a subset. And so my intuition is that's probably the case with some of these

00:32:29.440 procedures like meniscal repairs, which still offers no help to you or I right now if we're having knee

00:32:36.380 pain with an MRI that shows a meniscal tear. I actually don't know the answer in that setting.

00:32:41.000 That's interesting on so many levels. The first is what you're talking about with absorption

00:32:44.760 is this sort of lesson that there's huge individual variation in that stuff.

00:32:49.240 Staggering variation, by the way. I measure absorption synthesis in every single patient,

00:32:53.640 non-negotiable, no questions asked. And I am constantly amazed at how much variation exists.

00:32:59.440 Basically, three variables are determining this, right? It's sort of how much do you make,

00:33:03.140 how much do you absorb, and how much do you clear out of circulation with the LDL receptor?

00:33:07.440 And the variation is, it's overwhelming. And yet it's amazing to me that our profession looks at

00:33:13.360 just one metric, which is how much LDL cholesterol is there. And that's going to be the basis for

00:33:17.260 treatment. It strikes me as flying without instruments and deciding you only get to look at the horizon.

00:33:23.060 That's interesting. That gets at two things I want to remember. First, this idea of the

00:33:26.320 McNamara fallacy, you've heard of named after the Secretary of Defense during Vietnam, which is,

00:33:30.400 he said, are we winning the war or are we losing? Let's use something measurable,

00:33:33.320 our bodies versus their bodies. And since we're always winning, by that metric said,

00:33:37.560 okay, we're winning, obviously ignoring a lot of other important things.

00:33:40.260 What's the collateral damage? And yeah.

00:33:41.880 Yeah. So it's like, we often deem things important because they're easily measuring them

00:33:45.460 because they're important. But that individual variation gets to another thing that got me

00:33:48.660 interested that sort of caused the sports team to be very different from my proposal, which was

00:33:52.900 underlying the 10,000 hours rule, which is actually called the deliberate practice framework.

00:33:58.340 Because again, Erickson would not call it the 10,000 hours rule. There's something called the

00:34:02.560 monotonic benefits assumption. And essentially, if you have two people who've never done something,

00:34:07.480 for every equal unit of practice, they should progress exactly the same amount.

00:34:11.760 So it's that everyone's practice response is the same, is one of the assumptions underlying it.

00:34:16.440 And I started- How can that be?

00:34:17.940 It's not.

00:34:18.760 To say it's not would be an interesting conclusion or observation. But how could that even be the

00:34:22.360 null hypothesis? It seems so counterintuitive.

00:34:24.740 I don't know. Does it? It might be if people started from zero. I'm not sure what I would

00:34:27.400 think if they would progress exactly the same or not. But then people have done those studies,

00:34:30.580 you know- Meaning if we took a hundred people who have never spoken Spanish and we gave them

00:34:34.840 Spanish lessons, give me the evidence that if that were the case, wouldn't we see much more

00:34:41.120 homogeneity in schools?

00:34:42.960 I would think. I mean, Erickson would make the argument that, well, some of those kids

00:34:48.100 are engaged and some of them aren't, or maybe some of them had more practice before. So I think

00:34:51.740 to really evaluate this, you have to get some skill that nobody else has tried before, that

00:34:55.560 these people haven't done at all. Because who knows what they bring to school, all sorts of

00:34:58.440 other stuff. But there are studies like that where people who are sedentary do the exact same

00:35:02.120 exercise. One of the famous ones called the Heritage Family Study, where every member of two

00:35:05.740 generations of 98 families is totally sedentary, put on six months of identical cycling

00:35:10.780 training. And the range of variation was like a thousand percent, doing identical training.

00:35:17.420 Identical training. And you can see things like the military does this and people learning

00:35:20.740 sort of perceptual motor skills for air traffic controller simulations. And at very simple

00:35:25.540 simulations, actually, it's kind of like that. Like they converge if it's very simple. You just

00:35:30.340 have to see that one plane's coming, move one off the runway, and then it becomes about

00:35:33.860 how fast can you basically just move the mouse when you're doing a simulation. But as it gets

00:35:37.860 more complex, people start diverging with more practice. And so that monotonic benefits

00:35:43.000 assumption, I could find no evidence of it. It's like never shown up in a study of anything

00:35:47.760 unless it's an extremely simple task that everyone masters very, very quickly. And so again, I

00:35:51.960 was sort of saying that average is just obscuring individual variation.

00:35:56.540 So I keep preventing you from talking about this because I can't stop asking all these other

00:35:59.860 questions. But when you look at the title of the sports gene, the assumption would be, oh,

00:36:04.660 this is a book that explores the notion that a great athlete is genetically gifted. Michael

00:36:09.480 Jordan is Michael Jordan because he clearly has a set of genes that separate him from the

00:36:13.720 rest of us. And that's probably a bad example because it's so extreme. But talk to me about

00:36:18.260 some of the things that you found in that book that surprised you. And there were certain elements

00:36:22.780 of that book that didn't get that much attention, by the way, that I, in retrospect, thought I'm

00:36:27.060 surprised more people didn't fixate on that thing. Like, I don't know. Did it surprise you

00:36:30.340 what people drew out of that book as the most important insights?

00:36:33.760 Yeah. Yeah. I didn't think the thing that people were going to find the most controversial was the

00:36:36.960 10,000 hour rule, to be quite honest. I guess I didn't realize how...

00:36:40.720 How ingrained that is in our psyche.

00:36:42.400 Yeah. And that people were actually planning certain training plans, soccer teams, to 10,000

00:36:47.640 hours on the dot. And the thing that was the most important to me, I tried to write a book about

00:36:51.840 one of my closest friends and former training partner dropped dead at the end of a mile race. He was

00:36:57.300 like one of the top ranked guys in his age group in the country. Young Jamaican guy was going to be 0.65

00:37:01.280 the first in his family to go to college, all these things. And that kind of threw me for a loop. And

00:37:04.300 anyway, I got his family to sign a waiver allowing me to gather up his medical records. And it turned

00:37:08.800 out he had hypertrophic cardiomyopathy, textbook case, misdiagnosed because not easy to diagnose.

00:37:13.580 If he'd had a good family history, it turned out he probably clearly had a relative who... It's like

00:37:18.160 I started going to meetings for families that think they have HCM in their family. And they'd say,

00:37:22.740 well, we're not really sure, but cousin Jimmy died in the pool and he was a varsity swimmer.

00:37:26.800 And like Uncle Fred was in a one car accident. You're like, all right, these might be cardiac

00:37:31.040 arrest. And I wanted to write a book. One of the main reasons I got off the science track is I

00:37:35.320 wanted to write about sudden cardiac death in athletes. And that's what I tried to pitch a

00:37:39.080 first book on, but I didn't have the professional capital at the time and couldn't sell it. But

00:37:42.800 there's a section of that that I smuggled into one of the chapters in the sports gene, which is the

00:37:46.320 most personally important thing to me that I don't think I got asked about one time ever. So

00:37:50.500 I don't know if that's surprising or not. It just is. The most surprising findings of the book to me

00:37:55.340 were things like, I still don't know how to summarize the book, but that things that I

00:38:00.340 assumed were genetic, like the reflexes it takes to hit a hundred mile per hour fastball turn out

00:38:04.740 not to be the fact that major league baseball players don't have faster reflexes. That was

00:38:08.360 a surprise to me. I had assumed they do. The same is true by the way, for formula one drivers.

00:38:12.200 Is it? Yeah. So it turns out that. Oh, that makes sense. Perceptual motor skill,

00:38:16.020 right? They're probably using cues, like the changing size of something in there.

00:38:19.340 Well, one of my favorite exercises that you can see when you compare. So if you put a novice next

00:38:25.260 to Lewis Hamilton in a simulator, even adjusting for the speed at which things are moving, you won't

00:38:31.140 be blown away at where Lewis's eyes are at every moment that he is driving, how far ahead he is

00:38:38.680 able to see what's happening. So a few weeks ago, I was on the track and in an effort to really force

00:38:44.640 this type of learning. We have one camera that is actually looking directly at me, one camera that

00:38:50.180 is looking directly at the road and capturing all of the telemetry. And then I sit with my coach and

00:38:54.540 we review these two side by side because what I'm working very hard to overcome is the desire to

00:39:03.280 narrow my field to where I'm driving. And when you're going fast, that's innate. You don't want to

00:39:08.160 be looking somewhere way down the road. You're worried about falling off the road right now, but you can't

00:39:13.440 do that. So that's the thing that they've been able to train to do. It's not that they're going

00:39:18.040 faster. Sorry, that's what they have faster reflexes, which again, remember when I first was

00:39:22.700 shown these data, I was like, wow, looking at Senna, you'd think he has the fastest reflexes on the

00:39:27.620 planet. Right, right. Once you start reading about perceptual motor skills, it makes perfect sense

00:39:31.760 because any activity that's happening too fast, I mean, the things they have to do are too fast for

00:39:35.740 any human, even if they did have, were the top 0.001% of human reflexes, it wouldn't be fast enough.

00:39:41.380 So in boxing, there was this study I came across doing this sports gene where these

00:39:45.380 someone, I don't even know if they were doctors or scientists, did some test of Muhammad Ali.

00:39:50.380 I think that what they were trying to show was that even this brilliant black man has slower than 1.00

00:39:56.260 normal processing speed in his brain or something. And so they reported his, they would have him like 0.79

00:40:01.340 throwing a punch in response to a light or something like that. And they were saying like,

00:40:04.240 look, it's lower than average. And then, then someone said that they were testing it wrong. If you

00:40:08.020 subtract the delay for whatever cue they were giving, he actually, from first perceptible

00:40:12.340 motion to full extension, it was like 150 milliseconds, which is extremely fast. But

00:40:17.360 that means when he's throwing a punch also, I think other people throw punches that fast. I

00:40:21.080 don't think it was like an alone outlier for that, that it's faster than the minimum human reaction

00:40:25.460 time, which is a fifth of a second, just to see that something's in front of you and for that

00:40:29.840 message to get to your muscles, not to dodge. And so you literally have to be seeing things

00:40:35.680 before they happen or else you'd get hit by every punch. Of course, his genius at disguising what

00:40:40.040 he was doing was an attempt to confound people's ability to see the future. And so anything I think

00:40:44.460 that's happening at that speed, those aren't skills that anyone comes with. There might be

00:40:48.260 things that facilitate you downloading that software, but it doesn't, doesn't come with the

00:40:52.480 machine. You talk about kind versus wicked. I'm jumping between these books. I think we're just

00:40:57.440 going to end up doing that by the way. I really wanted to talk about the gene, but now I can't

00:41:00.560 stop, but help moving. So maybe use that example in boxing as an example. Is boxing a kind sport or

00:41:08.060 is it a wicked sport? And explain what those two distinctions mean.

00:41:11.080 Yeah. Kind and wicked are, so those are terms coined by the psychologist, Robin Hogarth in a kind

00:41:16.960 learning environment. He was trying to reconcile this issue in psychology and the study of expertise

00:41:22.980 about why some people who studied experts saw them get better and better and better with very narrow

00:41:27.580 experience. And some people saw them not get better, sometimes get worse or get more confident

00:41:31.820 and not get better. Like what was the difference? And it turns out that the difference often has a

00:41:35.860 lot to do with one, the way they're training, but also the environment that they are training in.

00:41:39.800 And a kind learning environment is one where all your information is clear. The next steps and goals

00:41:44.560 are totally clear. Work tomorrow will look like work yesterday. Patterns recur. And whenever you do

00:41:50.020 something, you get feedback that is immediate and fully accurate. On the other end of the spectrum.

00:41:55.120 So golf golf is a really kind learning environment because the ball is never actually moving towards

00:42:02.660 you. You're always starting with a static ball and there are almost a finite number of things that

00:42:07.600 you can see in that position and there's no rush. And you get automatic and real time feedback every

00:42:13.580 time something happens. So I think some of the people who study golf characterize it as like almost

00:42:17.420 an industrial task in the sense that part of what you're doing is trying to do a similar things over

00:42:21.740 and over with as little deviation as possible. Archery, which is my obsession, a very kind

00:42:26.280 learning environment. Absolutely. Totally kind. Absolutely. And a wicked learning environment on the

00:42:31.560 other hand is you might not know exactly what you're supposed to do next. You might not even know the

00:42:35.160 goal. Human behavior might be involved. There may be time pressure and work next year might not look

00:42:41.100 like work last year. And importantly, you don't always get automatic feedback. And sometimes when you get

00:42:45.840 feedback, it's delayed and sometimes it's inaccurate. One of the, there's actually a medical example that

00:42:51.320 I know the story you're going to tell. Yeah. Which is this doctor, this New York doctor who got wealthy

00:42:56.220 and famous because he could miraculously by palpating patients' tongues or feeling around their tongue with

00:43:01.340 his hands before they showed any symptoms, he could predict they would get typhoid. And he was right over

00:43:06.060 and over and over again. And one of his colleagues later observed using only his hands, he was a more

00:43:10.040 productive carrier of typhoid than even typhoid Mary. So he was giving people typhoid by touching their

00:43:14.600 tongues and getting the feedback that he was an amazing predictor and so would do it over. So he

00:43:19.780 was, the feedback was reinforcing the wrong lesson. So I wouldn't say most of us are in that wicked of

00:43:24.460 a situation either. But what Hogarth was doing was setting up this spectrum of learning. What do you

00:43:30.340 think is a bigger wickedness within the wicked environment? Because there are really at least two

00:43:35.460 variables that I think make that type of learning environment challenging. The first is the number of

00:43:43.060 scenarios you can face and the unpredictability of them. So in archery, the goal of archery actually

00:43:48.100 is to make every single shot identical, non-negotiable. So everything from the way you

00:43:53.840 stand to the way your shoulder sits to the way the release sits, we pay tremendous attention to

00:44:00.000 the feeling of the string on the nose and the feeling of the string on the corner of the mouth. I mean,

00:44:04.880 you're trying to reproduce the same thing ever and ever and ever. So part of it is, well, in tennis,

00:44:08.640 for example, there are an infinite number of ways that you could be standing and your opponent could

00:44:13.060 be standing and the ball could be coming with this spin versus that spin, or maybe not infinite,

00:44:17.220 but there are so many more variables. The second piece is this delay between feedback and reality,

00:44:23.120 which anybody who's ever tried to talk when they can't hear themselves or when there's a delay

00:44:27.680 realizes how much feedback matters. Which of those two do you think is more important in creating that

00:44:34.860 environment? I think delayed feedback is usually a killer. I think that the changing scenarios

00:44:40.820 is easier to accommodate with broader training in some ways. So this classic psychology finding

00:44:46.600 that can be summarized as breadth of training predicts breadth of transfer. Transfer is, and by

00:44:51.920 the way, I'd say tennis is definitely more on the wicked end than golf, but I would still-

00:44:55.560 It's still not, yeah.

00:44:56.560 As Hogarth said, most of us in the knowledge economy are playing Martian tennis. You see some people playing,

00:45:00.940 nobody's told you the rules, you have to deduce them. And by the way, they can change without notice.

00:45:05.240 And so breadth of training predicts breadth of transfer. Transfer is the term psychologists

00:45:08.780 use to mean your ability to take skills and knowledge and apply them to a situation you

00:45:13.100 haven't quite seen before. It might be similar, but something's a little different. And what

00:45:17.260 predicts your ability to do that is how broad your training was. If your training is broad,

00:45:20.700 it forces you, instead of expecting the same thing over and over, to build these sort of flexible

00:45:24.680 conceptual frameworks that you can bend when you see a new problem instead of just doing the same

00:45:29.320 thing over and over. So I think you can mitigate that with the right kind of training. The delayed feedback,

00:45:34.160 that really screws people up. In studies where people have to sort of drive remote-controlled

00:45:39.100 things, if they build in a delay between what they do and the movement, it completely screws them up.

00:45:43.660 All the way to, there's some interesting studies of software project managers. There's this famous

00:45:48.100 essay. I had to cut like 20 or 30,000 words from the book. This was something I had in there called

00:45:52.580 The Mythical Man Month. And it's an essay by this guy, Fred Brooks, who was like head of research at

00:45:58.460 Microsoft. And he went on to found the computer science department at the University of North

00:46:01.880 Carolina. And what he meant by The Mythical Man Month was he had noticed that when project managers,

00:46:08.220 when their projects got behind in software, if they were complicated, they would start adding more

00:46:13.260 person power, adding more man or woman power to the team. And that would cause the project to become 0.83

00:46:19.940 more late. And so Brooks's law is if you add people to an already late software project,

00:46:25.420 it will become more late in proportion to like how many people you add. And that's because there

00:46:29.140 was a delay between those people adding to the team. They needed to be assimilated. And the managers

00:46:33.820 never learned that lesson because of the delay. And so they keep doing the same thing over and over

00:46:37.900 and over. And a couple of researchers sort of followed up on that more recently and called this

00:46:41.600 the experience trap where these project managers, they come up with simpler projects where adding people

00:46:46.780 does help it get done faster because they can right away figure out what to do. And then they get promoted

00:46:52.300 and promoted and end up with more complicated projects. And in those cases, they do the same

00:46:56.500 thing and bring people on and they never learn about the assimilation delay that it takes. So

00:47:00.720 these researchers were saying we need to start telling them this is the time between you bringing

00:47:05.060 someone on and them making a positive impact. But they never learned that lesson because of the

00:47:08.940 feedback delay. So I think that's from the motor skills up to these much more sort of management

00:47:14.780 kind of softer skills, the feedback delay is really difficult.

00:47:17.660 That concept, ever since I read about it in your writing, I sort of look at the world a

00:47:22.420 bit differently now. I actually think of that question specifically. I'm like, how kind is

00:47:26.160 this? How wicked is this right now?

00:47:27.260 Me too. That's what happened. When I was reading this, I was like, oh, this is going to be the

00:47:30.120 frame that I'm going to think through, I bet, for everything in this book.

00:47:32.920 Well, I think about it a lot with kids. You have a kid, right? You've got one?

00:47:36.380 One, seven months.

00:47:37.420 Yeah. So, I mean, think about the learning that's taking place. Think about the neuroplasticity of a

00:47:43.060 seven-month-old. And for example, it's why, sort of going back to the example before,

00:47:47.980 if a child is deaf, it's going to delay speaking, not because they can't speak,

00:47:51.960 but because they can't get that real-time assimilation and feedback. And watching,

00:47:57.740 it almost makes you think about how much do you want to intervene when they're doing something

00:48:01.980 wrong too. I don't know if you've found that, but it's like, okay, as long as they can't really

00:48:06.140 hurt themselves, I should probably let them do that thing that is going to hurt, but hopefully not

00:48:11.380 irreversibly hurt. Yeah. Yeah. I mean, that's a huge question. Again, like I said, what I think

00:48:16.300 of as, this would make a terrible subtitle, so it's not the subtitle of my book range, is

00:48:20.040 the things that you can do that seem the best in the moment maybe are not the best for long-term

00:48:24.720 development. And I think that applies to parenting. Maybe we should talk about that

00:48:28.260 college admission scandal or something. The snowplow parent isn't the new, but clearly that's not the

00:48:33.400 best for long-term development in many cases. There's a story you write about, I think it's at the

00:48:38.520 Air Force Academy that illustrates that point. Explain that insight. This was one of my favorite

00:48:43.660 studies in range, partly because the experimental setup is so cool. You could only do this at the

00:48:48.780 U.S. Air Force Academy. It's a true experiment. It has randomization, it's prospective, and there's

00:48:53.760 blinding. Check all the boxes. Yeah. Yeah. So at the U.S. Air Force Academy, about a thousand students

00:48:58.440 come in every year, and they have to take a sequence of three math courses, calculus one, two, and third

00:49:03.580 course. And they are randomized to professors in year one, re-randomized in year two, and re-randomized

00:49:09.460 in year three. And the characteristics they come in with are spread evenly across classes. And so

00:49:15.480 para-scientists wanted to see, okay, this is a great experimental setup for looking at what is the

00:49:20.720 impact of different math teachers. So they followed about a hundred professors and 10,000 students over

00:49:26.380 a decade. And one of their main findings was that the better a professor was, oh, and everyone takes

00:49:34.160 the same test in every class also, and it's graded by committee, so there's no one can make their own

00:49:38.240 students do better. What they found was the better a professor was at getting their students to

00:49:42.100 overperform in calculus one, the more those students then underperformed in the next two follow-on

00:49:47.720 courses. So there's an inverse relationship between how well students with a certain professor did in

00:49:52.520 the first course and how well they did in the second and third. So for example, and how they

00:49:56.580 rated their teachers. So I think the professor who was rated the sixth best by his students in calculus

00:50:02.520 one, and his students got the seventh best scores overall, I think, out of a hundred professors,

00:50:09.400 his students did the seventh best, was dead last in how his students then did in the next two courses.

00:50:14.480 And essentially it turned out-

00:50:15.300 And just to be clear, were they dead last or were they dead last in improvement?

00:50:20.000 So there was a value-added score, which said, here are the characteristics these kids come in

00:50:24.740 with, and here's how we'd expect them to do. Are they over or underperforming?

00:50:28.420 Yeah. So it's not that they were the worst in Calc 2, it's that they underperformed relative to

00:50:34.000 where they came in.

00:50:34.740 That's right.

00:50:35.340 So they grew the least, maybe is the right way to think of this.

00:50:38.000 That's right. Yeah, absolutely. Compared to other people who came, other students who came in with

00:50:41.620 the exact same characteristics, they did worse than them in the follow-on courses. And what these

00:50:45.920 scientists found was that the way to get students to do the best in calculus one was to teach a

00:50:49.800 very narrow curriculum that was tailored to the test, where they learn a lot of what's called

00:50:53.800 using procedures knowledge, where they just learn how to execute algorithms and things like that

00:50:57.660 over and over. Whereas the professors who got rated worse by their students, and their students

00:51:04.400 did worse on the calculus one test, they learned more, it's called making connections knowledge,

00:51:08.700 where you have to draw together concepts essentially, and you're facing different types of problems

00:51:12.500 instead of repeating the same type over and over, which is another thing we should talk about

00:51:15.700 after this. But then when they go to the next courses, they have this more conceptual,

00:51:21.060 flexible knowledge. They're learning how to match a strategy to a type of problem instead of just

00:51:25.140 how to execute procedures. So they do better later on. And so there's this real conflict between how

00:51:30.340 they feel they're doing early and how they rate their professors and how they're really being set up

00:51:34.300 for later success, which is kind of wild.

00:51:36.860 And that spoke to me on many levels, but one of which being is I love mathematics. And I think we spoke

00:51:43.180 before and maybe we'll tell the story, get into that, get into the issue later about, I've always

00:51:47.000 wondered about the transition I made when I decided to take school seriously. But that teacher who I

00:51:53.020 think sort of turned my life around would go on to teach me calculus as well. And he had a very unique

00:51:59.440 style of teaching, which was you approach every problem through the lens of understanding what is

00:52:04.520 being asked physically and seeing if in the end, the question says, okay, Johnny throws a ball at this

00:52:10.760 speed in this angle. Where does it land? That might be the question. I mean, that's a very simple

00:52:14.540 calculus question, but that's an example. But as the problems get more and more challenging, he would

00:52:19.080 still really insist that you try to understand graphically using sort of functional calculus,

00:52:26.620 like graphically what is happening, algebraically what is happening, and numerically what is happening.

00:52:31.440 And how can you converge the numerical solution with the algebraic solution with the graphical

00:52:35.780 solution? He said in calculus, you should almost always be able to come up with an

00:52:39.660 estimate of where the answer is based on graphing the functions and looking at how they behave.

00:52:46.160 Hard. I mean, a lot of kids didn't do very well in calculus. And yet those lessons took me all the

00:52:53.380 way through honors math and engineering. And another sort of example on that path that really spoke to me

00:53:00.000 was in my freshman year of calculus, I met this guy studying in the stacks. His name was JP. And he

00:53:06.380 became a legend to me because he had simultaneously, he couldn't decide if he wanted to be a mechanical

00:53:11.420 engineer or an electrical engineer. So he did both. And he literally got both degrees in four years.

00:53:16.160 It's not impossible to get both of those degrees in five or six years, but to do both in four years

00:53:20.020 is crazy. And he said, the only way he was able to do it was he never wanted to memorize how a type

00:53:27.260 of problem was solved. He wanted to derive everything from first principles. At the time, I think we were

00:53:31.600 learning about Coriolis acceleration, which is basically the acceleration of a body in rotation

00:53:36.760 where the radius is changing. So named after a guy who failed to figure this out when miners in a shaft

00:53:43.080 were moving down. So now you actually had a shrinking radius relative to the center of the earth. And that

00:53:48.300 changes the forces on the elevator shaft. And he said, you realize you could derive that all from

00:53:55.480 Newton's first law. And I was like, I never thought of it that way. And he goes, yeah, let's go through it.

00:54:00.040 And it's about a one page derivation, but you can do it. And that lesson stuck with me for the

00:54:04.460 remainder of my life, which was, oh my God, if you just think of it in these broader, initially more

00:54:10.540 painful ways, it yields huge dividends. This gets at so many things. I'm not even sure where to start

00:54:16.820 in range. But the first thing that popped in my head was I just saw somebody tweeting research on

00:54:20.040 Twitter today about how active learning students actually learn more, but they rate themselves as

00:54:23.880 having learned less. And they also rate their teachers worse. So it turns out we're early on, we're not

00:54:28.340 actually that good evaluators of how we're doing because the feeling of fluency and learning makes

00:54:32.520 us think we're doing well, but we're actually not. This is what I wrote about called desirable

00:54:35.920 difficulties. But what you were talking about, I mean, that's serious making connections knowledge.

00:54:40.440 In Japan, they actually have a term called bansho that means the type of writing on the blackboard

00:54:47.680 that tracks all these different approaches to the same problem, sort of what it sounds like your

00:54:51.780 teacher was doing. But I want to get to the derivation point because there was some research

00:54:57.720 I wrote only a little bit about in range, but that I read a lot more about that was about what

00:55:01.460 college students understand about math, essentially. And there were some really startling examples where

00:55:07.860 one of the problems was like, I can't remember what the exact numbers were, but let's say it was

00:55:14.020 500 plus 200 equals 700. And the students were asked, how can you check if this was right? And so I'd say,

00:55:19.580 700 minus 200 equals 500. That's right. What's another way you can check if it's right? And they

00:55:25.440 wouldn't come up with 700 minus 500 equals 200 because they were taught to subtract the number

00:55:30.060 on the right of the addition sign. And when their professors were shown this sort of stuff. So this

00:55:34.540 is wicked feedback. So you get the feedback that the student understands because they know one way

00:55:37.880 to do it, but they actually don't understand it all. And the professors would say, oh my gosh.

00:55:43.940 And the students would, in interviews, they'd say like math is a system of rules and executed

00:55:47.680 procedures. Professors would say like, I went into math because I didn't have to memorize stuff

00:55:51.900 because you could derive it and it makes sense. It's just concepts. And so they were just in a

00:55:55.900 totally different place than the students were. And it sounds like you had some opportunity to

00:56:00.660 avoid that, but I think that's the norm of how math teaching works.

00:56:03.660 Yeah. And it's hard because my daughter who's in sixth grade, I'm doing this thing and it's really

00:56:08.340 hard. I mean, I can't imagine I'm the only parent that struggles with this, but I want her to love

00:56:12.380 math. I don't want her to view it as a subject. I want her to view it as math is more fun than

00:56:18.600 playing video games. Math is the most beautiful thing in the world. I want her to look out the

00:56:24.580 window and see math and see that math is a beautiful tool that we have to explain the world around us

00:56:31.520 through this thing. And yet every time as a parent, I try to ask her a question to ask her to think a

00:56:37.760 certain way. I end up sort of putting her on the spot. And so I'm struggling with this thing of

00:56:41.760 one, is it just that I'm a bad parent and I don't know how to do this correctly?

00:56:46.980 Definitely that you're a bad parent. I didn't want to say anything, but since you brought it up.

00:56:49.880 Or is it that a child needs a certain base of facts? Like they have to know a handful of things

00:56:56.980 and be confident with the language, like the times tables and all these other things before you can

00:57:02.080 even get them to start thinking beyond the problem. Like I think a lot about this actually,

00:57:07.260 because I feel like I'm underperforming on this actually.

00:57:10.460 It's a good point though. And I should say, so again, in research, these are called using

00:57:13.620 procedures knowledge, which is kind of the knowing stuff and making connections knowledge,

00:57:16.600 which is derivation, understanding concepts. And they're both important. It's just that like in

00:57:21.540 some of these famous studies in the United States, almost a hundred percent, sometimes literally a

00:57:26.180 hundred percent in classrooms would be the using procedures knowledge and not the making

00:57:29.920 connections. But I think that is a tricky thing. There's a study that came out after a range,

00:57:34.840 so I couldn't include it, but was on the topic of desirable difficulties. And it was about

00:57:41.160 interleaving, which is, well, I'll explain the study. Seventh grade math classrooms were randomly

00:57:46.320 assigned to different types of math teaching. Some of them got what's called blocked practice,

00:57:50.720 where you get problem type A, A, A, A, A, B, B, B, B, B, and so on. And the students get better

00:57:55.840 really quickly and they rate their teachers well and they rate their own learning well. Other classrooms got

00:58:00.320 interleaved practice where it's like as if you threw all the problem types in a hat and you pick

00:58:04.040 out randomly. And in that condition, the students are frustrated at first. They rate their own

00:58:08.360 learning low. They rate their teachers poorly. But again, they're learning how to match a strategy to

00:58:13.100 a type of problem instead of just do the same thing over and over. And come test time, they all took the

00:58:16.540 same test. The interleaving students who had interleaved practice blew the blocked practice

00:58:21.560 students away. It was, I think, like the largest effect size I've ever seen in an education study

00:58:25.680 that was randomized, 0.83 standard deviations. It was like taking a kid from the 50th percentile

00:58:30.840 and moving them to the 80th percentile. But they didn't like it early on. They don't feel like

00:58:35.260 they're learning. And so I'm not sure what the balance is as a parent where you know some of

00:58:39.720 this desirable difficulty is in the long-term desirable, but you also don't want to turn someone

00:58:44.320 off from the subject. So what's that delicate balance? I think that's sort of kind of the art of

00:58:49.640 coaching in everything we do, whether that's someone in sport trying to develop someone for the long-term

00:58:53.740 or a parent is like, how do you balance maintaining enthusiasm with optimal development and helping

00:59:00.060 someone have that vision of their future self, like your professor did for you without having

00:59:04.960 them be burned out? I don't know. I really think that's why there's what like great coaches kind

00:59:09.720 of do is they figure out how to balance these things, when to make things difficult and when to

00:59:14.380 allow things to be easy and sort of more easily inspirational. But I think that's an art as well

00:59:18.660 as a science. So what was sort of the takeaway of the idea that, and I don't like using the extreme

00:59:27.080 examples because they're sort of silly and that's the problem with them. But if you look at a high

00:59:33.240 school track team, okay, and you look at, you pick a big division one school and you look at their

00:59:39.600 sprinters, are they genetically predisposed to be sprinters, but without a certain degree of training

00:59:48.600 could never appreciate it? Or could a certain amount of training overcome a lack of genetic

00:59:53.960 predisposition? I mean, how do you feel about that today versus when you wrote the book versus when you

00:59:59.660 wrote the article? Because again, you're getting smarter as we go. Yeah. When I wrote the article,

01:00:04.000 I was more convinced than I had been before the article that genes were unimportant, completely

01:00:10.720 unimportant. And by the time I wrote the book, I came to feel that there were sort of two extreme

01:00:15.800 camps, one that felt genes have no influence on performance. And another that I'd say the other

01:00:22.440 extreme wasn't that practice has no influence. I don't think anybody thought that's uncontroversial.

01:00:28.500 So I think one extreme was only practice matters. And the other extreme was practice and genes

01:00:33.660 matter. I think that is not as extreme saying they both matter. Which, and my bias is the latter.

01:00:38.800 My bias is just that. I mean, that's what the evidence shows. Yeah. And it's like people don't

01:00:43.660 randomly pick to train for the marathon to the hundred meters if they're trying to get to the top

01:00:47.240 level because there's some zero sum physiology going on there. And so I think to be a sprinter,

01:00:53.180 you're not turning a cart horse into a racehorse as the saying goes. Like you have to have some

01:00:57.180 predisposition to being fast and being explosive. And there's a reason why those people are particularly bad.

01:01:03.660 Usain Bolt would be a worse marathon runner than a random person picked off the street

01:01:07.200 because it's different physiology. But, but I also think it's important to note that at the

01:01:12.340 beginning of the season and six months later, someone like him are very different in how fast

01:01:16.300 they are. Even over the course of one season, they change how fast they are. So I think the

01:01:19.400 practice is, is incredibly important, but you also need talent.

01:01:23.680 Now, which is maybe a reasonable segue into a story that I know you just get asked about all the

01:01:28.100 time. So, but I also feel like for the sake of the listener, if they haven't heard you on

01:01:31.300 another podcast, it's worth them hearing you explain the difference between Roger Federer and

01:01:36.880 Tiger Woods. But of course I, once you're done with this, I want to kind of go down more extreme

01:01:40.680 examples and stuff. So is it safe to say that the contrast between Tiger and Roger are elite

01:01:47.200 examples of opposing views?

01:01:49.480 Yeah, yeah, definitely. And it's sort of interesting because, well, I'll give the quick versions of the

01:01:53.540 story first before I criticize myself. The Tiger Woods story, I think even if people don't know

01:01:58.040 the details, they probably kind of absorb the gist. Seven months old, father gives him a putter.

01:02:01.660 Not trying to turn him into a golfer, by the way. Oh, this reminds me, we should talk about how the

01:02:04.520 Tiger and Mozart stories I think are told wrong after this. So I'm just putting that on our cork

01:02:08.560 board here. Ten months, he starts imitating his father's swing. Two years old, you can go on

01:02:13.660 YouTube and see him on national television demonstrating his swing on the Mike Douglas show.

01:02:17.560 At three or four, he starts saying, I'm going to be the next Jack Nicklaus. Fast forward to age 21,

01:02:22.240 he's the greatest golfer in the world. And that's sort of the quintessential, I think that story has

01:02:25.840 seeped into culture so much that people who don't even follow golf ever kind of know it.

01:02:30.440 Federer, on the other hand, every bit as famous as an adult, but obscure.

01:02:35.060 Every bit as dominant.

01:02:36.060 Every bit as dominant, more so over a longer time.

01:02:38.440 For sure.

01:02:39.080 And when he was a kid, he played some basketball, badminton, tennis. His mother was a tennis coach,

01:02:44.640 refused to coach him because he wouldn't return balls normally. He continued on to do skateboarding,

01:02:49.020 swimming, wrestling, soccer. When his coaches wanted to move him up a level, he declined because he

01:02:54.680 wanted to talk about pro wrestling with his friends after practice. He went on to play handball. Maybe

01:02:59.520 I said volleyball already. I'm not sure. Some rugby.

01:03:01.960 Do we know how good he was at these other sports, by the way?

01:03:03.980 He was good at soccer, for sure. Most of the other sports, I'm not really sure. But he kept playing

01:03:08.420 badminton, basketball, soccer longer than some of the others. And then soccer was the one that he

01:03:13.260 finally decided he had to choose between soccer and tennis.

01:03:16.820 And what age is that when he's having to make that decision?

01:03:19.300 I think he was starting to think about that as he was entering his teens, basically.

01:03:22.180 And he wasn't focused. Tiger's saying, I'm gonna be the Jack Nicklaus when I'm four. Roger was

01:03:27.540 actually, when he first got good enough to get interviewed by a local paper, the reporter asked

01:03:33.220 him if he ever became a pro, what would he buy with his first hypothetical paycheck? And he says,

01:03:36.920 a Mercedes. His mother doesn't want him putting all his eggs in this basket. It's like appalled.

01:03:40.600 And I asked the reporter if she can hear the interview recording. And the reporter obliges. It turns

01:03:44.260 out Roger said Mercedes in Swiss Germany. He just wanted more CDs, not a Mercedes. And so his mom's like,

01:03:48.700 all right, fine. But it's just very different in every way from the Tiger story. And the way I sort

01:03:54.120 of used these is as a device to start range, first to set up just the concept going forward. The book

01:03:59.680 proposal was titled Roger versus Tiger. And it was going to be like, when should you be a Roger? And

01:04:02.940 when should you be a Tiger? But they felt everyone would think it was a biography of those two guys,

01:04:08.000 my publisher. And I thought it was telling that we only know, only hear one of those stories.

01:04:12.720 And my question was, which of these is the norm? Obviously, they both worked for these

01:04:16.240 individuals. And I think there's many paths to the top as there are people. But what was the norm?

01:04:20.260 And it turns out that the norm when scientists track athletes en route to becoming elite is that

01:04:25.100 early on, they tend to have what they call a sampling period, where they try a wide variety

01:04:28.420 of activities. That often includes things like martial arts and dance and doesn't just have to be

01:04:32.880 other sports, gymnastics. And they learn these broad general, these physical skills. And they learn

01:04:39.560 about their interests and their abilities, importantly, and delay specializing until later than peers who

01:04:43.880 plateau at lower levels. And that turns out to be the norm. And so I sort of felt we should let people

01:04:50.180 know what the norm is, instead of always focusing on these very few exceptions that happen to be in

01:04:55.720 like the most kind learning environment, sports, and that aren't that good to extrapolate to everything

01:04:59.720 else.

01:04:59.900 Now, how much of that do you think is the psychology of it and the neuromuscular physiology of it? So

01:05:09.580 looking at Roger, for example, who I know very little about Roger Federer, that's not the obvious

01:05:14.760 stuff that most people who are not, I'm not a tennis fan or anything like that. But part of his

01:05:19.240 longevity, is it possible to be explained by the fact that he never burnt out versus just sort of

01:05:25.200 slowly acquired a love for this thing versus had it shoved down his throat? I'm sure like there's lots

01:05:31.060 of stories of trying to create these tennis prodigies where it probably backfires because by the time the

01:05:35.900 kid is 16, they're great, but they've lost the desire. So it's more of a above the neck phenomenon

01:05:42.300 than a below the like environment. How much of it do you think is that versus in playing all those

01:05:47.420 other sports, Roger actually developed synaptic connections that served him better in tennis?

01:05:56.840 He basically created a bigger foundation across his neuromuscular system that ultimately came to

01:06:04.240 serve him when he specialized. So there are a couple points. So this will be a sort of a longer

01:06:08.680 point for us because you bring up a couple of good things. Initially, I thought that most of the

01:06:12.860 effect was going to be accounted for by the fact that if you allow people to delay selection, it's

01:06:17.520 more likely you get them in the sport that they're the best at. Whereas we know the earlier you force

01:06:21.940 selection, the more likely you put the wrong person in the wrong sport. And when selection occurs

01:06:25.340 really early, you end up seeing this huge relative age effect where coaches just pick for kids who are

01:06:29.180 born early in the selection year because they're seven or eight or nine or 10 months older than their

01:06:33.640 cohort. And so they're at young ages, that's a huge difference. And so coaches mistake biological

01:06:38.800 maturation for talent. So I thought most of it was just going to be the fact that if you delay

01:06:42.580 selection more, you'll get the right people in the right sport more. But then I started coming across

01:06:46.680 these studies of German national soccer players or people in the national development pipeline,

01:06:51.420 where they were matched for ability at a certain age, tracked for several years, and they see who

01:06:56.220 improves more. And at certain ages, you have to focus eventually, but at certain ages, like in the early and

01:07:00.860 mid-teen years, it was those who did a wider variety of activities. And so then I started to think maybe

01:07:05.580 there really is something to the skill benefit, which didn't surprise me intuitively, but to see

01:07:10.800 it empirically was interesting. And then I spent some time with, I should say, by the way, my colleague

01:07:15.740 John Wertheim asked Roger Federer about this, one of the questions you asked on the Tennis Channel

01:07:19.340 recently, and Roger Federer said it contributed to his not burning out. You know, you never know how much

01:07:23.440 to trust someone's own story, but it's worth noting that that's what he said. I spent some time with

01:07:28.200 this physiologist for Cirque du Soleil, and he noted that they started implementing this program

01:07:33.980 where, because some of the performers are former Olympians and things like that too, where they

01:07:38.600 would have performers learn the basics of three other performers' disciplines. Not because they

01:07:42.300 were ever going to perform them, but let's see if it would vary up what they were doing, maybe reduce

01:07:46.500 stress-related injuries and stuff like that. And they track their injuries next to Canadian 0.93

01:07:51.280 Gymnastics. I guess it's a Canadian company. And he said it reduced their injuries by like a third.

01:07:55.800 So they implemented it. They must feel really strongly about it if they're taking away from

01:08:00.400 practice time for those performers their main discipline. There seems to be something. This

01:08:04.340 showed up in another longitudinal study of young athletes where the best predictor of suffering

01:08:08.800 what they called an adult-style overuse injury was how specialized the athlete was. And it wasn't

01:08:15.380 necessarily their total time spent in physical activity. It was if it was just the same thing over

01:08:19.740 and over. So there was like some protective effect from diversifying. We can guess at what that is. I'm sure

01:08:24.580 your guesses would be better than mine. But ultimately, my feeling is the Roger pattern is

01:08:30.080 more prevalent. One, because you have that breadth of training that predicts breadth of transfer. You're

01:08:34.800 exposed to much more neuromuscular stuff, much more perceptual stuff. And when the challenge gets

01:08:39.440 harder as you go up, you need to draw on those. There's a funny book called Extraordinary Tennis for

01:08:45.520 the Ordinary Player. I think by this guy, Cy Ramo, who's better known as the father of intercontinental

01:08:50.440 ballistic missiles, but also wrote some books about tennis and a couple other things. And in it,

01:08:56.560 one of the interesting things in it was that he shows, he does some like serious analysis of

01:09:01.680 gameplay at different levels and shows that even for good amateur players, something like 80 or 90%

01:09:08.100 of the points are scored by just keeping the ball in play and someone making an error. And then when

01:09:12.080 you get to the elite level, it's totally exactly the opposite. It's like 80, 90% of points have to be

01:09:16.440 earned. And that completely changes the kind of game that you're seeing. So I think the challenge

01:09:21.460 that a lot of these athletes are facing really changes a lot as they go up in levels. And so

01:09:25.760 they really want to have that kind of breadth of training in this experience, responding to

01:09:30.320 different types of... Do you think this overlaps with the Air Force Academy example where having the

01:09:34.960 harder, more orthogonal education in calculus prepares you for more real world problem solving,

01:09:43.120 which is what's happening as you go from calc one to two to three?

01:09:45.720 I think so. I think so. And I mean, again, that's why I think the theme of the book is

01:09:49.060 the things that'll cause you to be the best today might not be the best for five or 10 years from

01:09:52.820 now, or the best way to develop a 10 year old might not be the best way to develop a 20 year old

01:09:56.800 or certainly isn't. And it's also, I think the specialization model may well, there's a surprising

01:10:01.900 dearth of research in golf for how popular sport it is. But I think the specialization model may well

01:10:07.260 work for golf because you're not facing some of that same stuff. It is a very kind learning environment.

01:10:12.100 And I mean, I guess the best guy in the world right now, Brooks Koepka didn't come until later.

01:10:16.080 And it's unclear if he even likes golf, but I could see, it makes sense to me that the

01:10:20.880 specialization model would work or at least not be deleterious in golf. Whereas in the other sports,

01:10:25.840 I'm...

01:10:26.320 What about the sports where physiology is undeniably huge? So the big three being swimming,

01:10:32.080 cycling, and running. Cycling is hard to talk about because people tend to conflate the use of

01:10:37.300 drugs with somehow discounting the remarkable physiology of these guys. But if you take a

01:10:42.420 Chris Froome, for example, four time Tour de France champion, grows up in Kenya, we can speak to what

01:10:48.240 the importance of early exposure to hypoxia could have been. What is our belief about the training

01:10:56.540 effect and the duration of the training effect necessary to produce world-class athletes at that

01:11:00.880 level? Because to me, that's as foreign... Like if you said to me, Peter, I'm going to put you in a 1.00

01:11:07.120 time capsule. You're going to be 16 or 14 years old again. Knowing everything you know today, if you

01:11:12.660 have to become a professional athlete, which sport would it be in? The answer is none. I positively know

01:11:18.440 there's nothing I could ever be good enough in. And that includes if I was willing to do everything

01:11:24.720 that was necessary to become the best cyclist, runner, swimmer, I just couldn't do it.

01:11:28.460 You'd have to do what some countries host the Olympics. What they focus on is just recruiting

01:11:32.460 more people in the much less competitive sports. If there's a basket weaving... I have a friend who

01:11:37.080 joked about this because my interests are so diverse. He goes, Peter, do you realize that if

01:11:41.580 someone comes up with an Olympic sport that requires solving a differential equation, driving a car fast,

01:11:48.300 shooting a bow and arrow, and doing a deadlift, you could be one of the best? That's basically

01:11:53.180 modern pentathlon. You're pretty much there. And he listed off 10 other really stupid esoteric

01:11:58.200 things I do. And I'm like, yep, that's great. That's my claim to fame. 0.99

01:12:01.380 Modern pentathlon is like fencing, horse riding, swimming, running. It's like, we should have that.

01:12:06.660 I think there was a time when chess and fiction writing were in the Olympics. So there's hope for

01:12:10.880 you. 0.98

01:12:11.980 I suck at both of those things. 0.99

01:12:13.800 Or like the Brits, I think in the winter sports when skeleton got introduced, which is where you 0.97

01:12:18.240 slide.

01:12:18.380 Yeah, yeah, yeah. Hilarious.

01:12:20.160 And there's a great innovation story about that if you want to get to that. But one of the guys who I

01:12:23.920 talked to in their program said, we've got this down like 80% to a science. We make like an open

01:12:27.540 call for women. We like do some measurements. We know what size they need to be. We know what kind 1.00

01:12:32.140 of explosion they need to have. We pick the gold medalist. And they've done that. So that stuff's

01:12:37.180 kind of amazing.

01:12:38.280 But do you think that for these super physiologic things where presumably mitochondrial density,

01:12:44.480 mitochondrial efficiency, fiber distribution matter the most, is that more of a golf? Is that more

01:12:51.080 tiger? Is that more Roger? Notwithstanding the psychological component, which is obviously enormous.

01:12:56.740 So I think in sports in general, more people, a hundred years ago, you could have come to the

01:13:01.200 Olympics and been the only person who knew anything about training or the only person who was really

01:13:04.280 talented and win. Now I think in every sport, many more people are ruled out by either their nature

01:13:09.300 or their nurture, no matter what the sport is. But I think in those-

01:13:12.340 Meaning they're not willing to train hard enough, no matter how much their ability. So that rules

01:13:15.680 them out.

01:13:15.780 Or they don't have the opportunities. I mean, a huge portion of people in the world don't have an

01:13:18.780 opportunity to be exposed to most sports. So they don't have the opportunity. I mean, I think one

01:13:21.820 thing, the impact of Title IX in the U.S. is showing in how dominant we are in some women's 1.00

01:13:26.540 sports. Where I think I just saw, I think Ross Tucker, he's at Science of Sport, a prominent

01:13:30.700 sports scientist. He tweeted that the United States has something like 40% of all the women's

01:13:35.660 registered soccer players in the world. So like, of course we're awesome because we're giving more

01:13:39.220 opportunities. So I think a lot of people, either they're not willing to train, they don't have

01:13:43.500 access to training, one of those, or they don't have the nature for it. And the more competitive

01:13:48.160 the sport is, the more important that is, right? Because obviously if we had everyone do identical

01:13:53.040 training, only genes would separate them. And if we had everyone be identical twins,

01:13:56.160 only training would separate them. But I think people are more quickly filtered out by their

01:14:00.880 nature and things like sprinting.

01:14:02.700 So how old are you?

01:14:04.220 I'm 38.

01:14:04.980 38. So I'm 10 years older than you, but directionally, like we're both clearly past our

01:14:09.200 prime, right? In the sense of like-

01:14:10.960 Sports-wise.

01:14:11.280 Sports-wise, yeah. I think I'm past my prime in everything, but clearly athletically. Is there

01:14:15.900 a reason that I, let's assume I'm a decent athlete, which I'm actually not, but let's assume

01:14:21.060 I was. Could I, if I decided tomorrow, like I want to play tennis, I don't think anybody would

01:14:28.000 ever assume I could become a good tennis player. But is it because the deliberate practice argument

01:14:33.060 would be there aren't enough hours left for you to devote to this? Another argument would be, no,

01:14:38.560 you've missed a critical window. Just as we say, by the way, this might be totally BS. I'd like to

01:14:42.740 hear your view on it. This view that we learn language is best at a certain window. And once

01:14:46.540 that window closes, it's sort of like a growth plate closing over a bone. It becomes really hard

01:14:51.200 to learn languages thereafter. Is there something about this critical window of exposure? I guess

01:14:56.340 is really the thing I'm trying to get at with this question when it comes to physical talent.

01:15:00.500 That's a tough question. And there's this book by a neurologist called Why Michael Couldn't Hit.

01:15:04.400 And it's about why-

01:15:05.220 I love this book. It was one of my favorite books.

01:15:07.620 So he was saying Michael Jordan kind of missed the critical window for developing the perceptual

01:15:11.940 anticipation skill that you need to see things that are coming because your reflexes are too slow.

01:15:16.900 Right. Why the greatest athlete we'd ever seen of a generation was a 188 hitter in triple A ball.

01:15:23.300 Yeah. I'm a little bit of a gadfly about that though, because I think that was a great book. I

01:15:26.560 loved the book.

01:15:27.260 I love the Wayne Gretzky story, by the way, but we'll come back to it.

01:15:29.520 Great. Yeah. And one season, I think Michael hit like 220 or something like that in minors,

01:15:33.500 which I think if we went down and picked a random person off the street, they would hit zero

01:15:36.660 in AA. These are people who are stars of college and high school teams or foreign teams. So did he

01:15:42.100 do well or poorly? I don't know. If he had hit 10, you know, I wouldn't have been surprised because

01:15:46.560 he hadn't been playing in a long time. So I'm kind of impressed with what he did. But I also think

01:15:49.900 there's something to-

01:15:51.440 I mean, I think the point though was the expectation. It's not that hitting 288 or whatever

01:15:55.440 in triple A ball is horrible. No, it's just, why isn't the guy that seemingly has the best

01:15:59.920 hand-eye coordination in the world immediately able to absorb it?

01:16:03.340 Right. And in that sense, I do think there was probably something to the critical period.

01:16:06.400 And there are always people, there are always exceptions to everything. One of the most dynamic

01:16:10.160 players in baseball now is Lorenzo Cain. And he did not play a game of baseball until age 16.

01:16:16.320 That surprises me. He did not know how to play. I think most people would need to have some

01:16:21.460 exposure, not necessarily specialized, but have some exposure at that age. So there's always exceptions.

01:16:25.380 But I do think you want some of that early exposure, partly just because you run out of time,

01:16:30.400 because you want your perceptual expertise to coincide with your physical, right? So you're

01:16:35.520 just under a serious time limit. And in relating that to language, I actually do think, and you

01:16:40.000 know, there was a guy who tried, dropped everything when he read about the 10,000 hour rule and decided

01:16:43.700 to try to become a pro golfer by doing 10,000 hours exactly. And he got Erickson to consult with him

01:16:48.480 and everything named Dan McLaughlin.

01:16:49.820 How far did he get?

01:16:50.500 He got to something like 7,000 hours or something like that. And then he, he stopped, he was having

01:16:56.600 injuries. And what happened was, I think he didn't make it and he didn't nearly make it,

01:17:02.460 but he got really good. So he got better than like 90% of amateurs or something, but wasn't nearly

01:17:07.380 going to get into Q school qualifying for a professional tour. And so what I think,

01:17:12.920 and I didn't know if he was going to make it or not, because my point has been like,

01:17:15.260 people were saying, Oh, you were right. He didn't make it. I wrote a little bit about him. I'm like,

01:17:17.920 that's not what I said. I said, there's huge individual variation. If he's going to make it,

01:17:21.260 it's not going to be at exactly 10,000 hours. And so he didn't make it.

01:17:25.340 By the way, do you think there's enough variation that it almost is uncoupled? Like what would be

01:17:28.680 your 90% confidence interval on that? Or does it even matter? I mean, I think there has to be a

01:17:36.260 90% confidence interval, right? If you think about it, is it a thousand to 40,000? If you had to say

01:17:41.500 90% of people that become, that achieve mastery do so in a certain amount of deliberate practice.

01:17:45.820 I think it really depends on the sport. I listed in the sports team, I listed some hours and it

01:17:49.680 varied a lot by sport. And so I think it sort of depends what it is. Most of them were in sports

01:17:54.480 were lower than 10,000 hours, significantly four to 6,000 hours kinds of things. And again,

01:17:58.320 these are averages in chess. It was higher than that. So I think it depends. I think it's sport

01:18:03.400 dependent. And again, something like skeleton, there's a great paper called ice novice to winter

01:18:07.720 Olympian in 14 months, where it's basically just pick somebody and then they can go to the Olympics.

01:18:11.540 So I think it depends on what you're doing. And also when I sat in on a Harvard Business School

01:18:16.120 class for some reporting I was doing for range, you just reminded me of something where the

01:18:21.000 professor asks the students, he asked them all these things like, how many Subway sandwich shops do

01:18:26.340 you think there are in America? Give your 90% confidence interval. And basically, we can't do it

01:18:31.960 because if you're asked that question, if you're asked 20 of those questions, you should be able to

01:18:36.260 get 18 correctly in a reference range. I've played this game myself and with people. I've never seen

01:18:41.920 anybody come close to it. People don't go big enough. Give your 90% confidence interval and

01:18:45.660 you'd be better off going like, well, I know there's 10, you know, to like a million. And

01:18:50.820 instead they go much narrower and they end up missing almost all of them. So what would my 90%

01:18:55.860 confidence interval be? And it's sort of, it depends what counts too, because there's some accounts

01:19:01.040 of athletes who have done a bunch of different sports. And so there are studies that show that

01:19:05.920 invasion sports are the ones that require anticipatory skills. People can't see me doing

01:19:11.120 air quotes. I guess we're on a podcast, but that's the term that they use in the invasion sports where

01:19:16.240 you have to anticipate things that are happening faster than you can react to. Boxing, soccer,

01:19:21.680 whatever, baseball, things are flying, trying to get past you. There's some studies that show that

01:19:26.480 people who have done a variety of invasion sports will then pick up any subsequent invasion sport

01:19:32.520 more rapidly. And so I think there was a case of one woman who had played a variety of sports and

01:19:39.520 then it only took her like 500 hours to become one of the best basketball players in the world,

01:19:43.100 but she'd played netball and she'd played a whole bunch of other sports, volleyball and all this

01:19:47.580 stuff. So it sort of depends. What do you count as deliberate? Like Erickson wouldn't count that as

01:19:51.180 deliberate practice because it's not the same sport, but clearly it is like lowering the threshold.

01:19:55.240 Well, and that's really the Federer point, isn't it? It's that all those other sports,

01:19:59.260 he's doing soccer, badminton, et cetera. I mean, they're still training in a more diffuse way,

01:20:05.420 a set of skills that obviously have gone on to serve him greatly in tennis.

01:20:08.440 Yeah. And I think this relates to language. I want to segue to language a little bit, which is

01:20:11.460 I wanted to write about language in the book. As I was going through all the research,

01:20:17.480 I found so much of it contradictory and confusing that I decided to kind of stay away from it largely

01:20:23.560 because I just couldn't figure out. I was hoping there'd be, in my proposal I wrote about this,

01:20:27.520 I'd seen this really cool study and I had video for it where infants who were being raised bilingual,

01:20:33.220 they were given like this lucite box sort of thing and plexiglass or whatever it was.

01:20:39.500 And there was some object and they had to find the opening in this clear box and get an object out

01:20:44.200 of it and then put some other object in it. And the ones who were bilingual would try more different 0.99

01:20:49.240 strategies. And the researchers were saying, well, they think differently and they have more

01:20:53.060 executive function. And I thought that was tantalizing. And I loved the video, but I just,

01:20:58.040 the research was all over. And my conclusion was kind of, there's a lot of tantalizing stuff,

01:21:01.700 but nobody's quite sure really for a lot of it. But one that I did think was pretty strong was the

01:21:06.480 idea that people who grew up bilingual had an advantage for then learning a third language without

01:21:12.380 being taught it formally. So there were studies where they'll be given like a fake made up language and fake

01:21:16.920 grammar and just have to learn it by immersion. And they seem to do that better. And I think that's

01:21:21.320 sort of akin to what we see in sports. And with regard to the sensitive period, I do think there

01:21:26.920 is a real sensitive period in language where I think about after like age 12, you're not going to make

01:21:32.360 something your native language anymore. And I think there are cases where kids, feral children cases,

01:21:37.560 these rare cases where a kid like grows up in the woods or isolated from people. And if they haven't

01:21:41.260 learned some, if they haven't had exposure to language by age 12, they never learn it basically.

01:21:45.240 And that also happens to be about the age you have to start. If you don't start studying chess

01:21:49.160 patterns by age 12, your chances of reaching international master status drop. International

01:21:53.280 master status, again, one down from grandmaster. They drop from like one in four to one in 55.

01:21:57.920 So I think there are some critical periods, but I don't think for most things, maybe for these feral

01:22:03.800 children, but for most things, I don't think it's nearly the expiration date that people think it is.

01:22:08.240 I think most people can get better at most things than they think they can.

01:22:12.020 And they can get better at most things than they think they can at older ages than they think they

01:22:16.780 can too. Yeah. It's funny you say that that's kind of where I wanted to pivot for a second,

01:22:20.740 which was all of this discussion is interesting through the lens of being the absolute best

01:22:25.180 tennis player or the absolute best golfer on the planet. But isn't that really besides the point

01:22:30.400 for most of us? Because if there's 7 billion of us on this planet, 6.9999999 billion of us are never

01:22:37.000 going to be good enough at anything to make a living at it outside of our day jobs. Anyway,

01:22:41.380 I was asked on a podcast recently why I love archery and driving a race car so much. And part of it is

01:22:50.840 there is still every six months, I'm still able to look back and appreciate the progress I've made.

01:22:57.940 In other words, I'm coming from a place of being so not expert at these things that the joy

01:23:05.400 is actually in the monotonic increase in skill. That is actually the joy to me. It's the, I don't

01:23:13.680 want to say mastery because that implies you are mastering it, but it's the path towards mastery that

01:23:19.220 is more joyous. And I don't know, it's almost like on some levels it must, maybe it's not that much fun

01:23:26.120 to be the best in the world at something because by definition, you only have one way to go at some

01:23:31.220 point. And that's probably a lot less enjoyable than working your way up the curve. And I guess

01:23:36.780 generalists have the ability to, I don't know, this gets more into the psychology of it again,

01:23:41.820 but you don't have to tether your identity to just this one thing. I always feel bad actually for

01:23:45.960 athletes who it's a brutal, brutal way to make a living in the sense that you have a far narrower

01:23:52.000 window in which you can be the best at something versus like say any sort of normal career.

01:23:57.020 Yeah. And if you interact with a lot of athletes, you hear a lot of them say,

01:23:59.100 now it's just a job and it was something that they loved at a certain point. I think we were

01:24:02.920 talking about Ayrton Senna earlier. I seem to recall him saying something like the times of

01:24:09.500 the sport that he loved the most were not when he was necessarily on the top of the world. It was

01:24:13.380 earlier. It was actually during his last two seasons in karting. It was racing this British

01:24:17.740 guy named Fullerton. Everything he did that karting was his favorite time, right? Absolutely. It was

01:24:23.300 pure bliss, no politics, as he said, just pure racing. I just actually interviewed a friend of

01:24:30.080 mine the other day. This podcast, it'll come out at some point in relation to this one. I'm not sure

01:24:34.160 when, but same question I asked her. She's an Olympian. And it was like, okay, you were the

01:24:39.160 second best in the world on this day. You got a silver medal. When was the sport the most fun? And

01:24:43.940 it's like, oh yeah, two years before I went to my first Olympics was when it was its most fun.

01:24:48.140 Yeah. Yeah. It's amazing. I mean, this makes me think of so many things. One, I think people that

01:24:53.860 are performers need other stuff to do. There was a study I mentioned in range where Nobel laureates

01:24:57.880 are 22 times more likely than a typical scientist to have like a serious aesthetic hobby, even though

01:25:01.940 they're certainly not as good at it as they are at their day job. For me, I noticed while I was

01:25:06.540 writing this book, somewhere along the line, I sort of started to forget what I'd loved about writing,

01:25:12.580 I think. And I keep this thing I call my little book of small experiments, where at least every other

01:25:17.320 month, I think of some skill I'd like to learn a little bit about, or some interest I'd like to

01:25:22.220 explore a little bit. And I put a hypothesis of, well, what could I try to, almost like my grad

01:25:26.720 student notebook, what could I try to get some insight into this? And it forces me to try something

01:25:31.860 new every other month. It doesn't have to be a big deal. Maybe it's some job that I don't know about,

01:25:35.520 and I just have to find somebody to talk to about it. But when I was sort of losing my enthusiasm for

01:25:42.160 the kind of writing I was doing and starting to feel more pressure, because my first book was a

01:25:45.660 surprise success. And then all of a sudden, you get a lot more pressure after that. I took an online

01:25:50.260 beginner's fiction writing course. I've been reading this book about the Zen concept of beginner's mind,

01:25:54.860 where you just always keep your mindset as a beginner, never as I've arrived. And in this class,

01:26:00.900 it's like, nobody cares who you are, nobody cares what you've done. You get back into that feeling of

01:26:05.820 being uncomfortable. And I loved it. I mean, it reminded me what I loved about doing this. It made me

01:26:10.720 totally uncomfortable, because it's still writing, but it's very, very different. I think it had these huge

01:26:15.120 benefits. One of the exercises was you had to write a story using no dialogue whatsoever. And

01:26:20.320 after I did that, I went back to my book manuscript that I had and was like, I've been leaning on quotes

01:26:24.040 in a lazy way when I don't understand stuff. If I don't understand it, the reader's not going to

01:26:27.060 understand it. So I need to learn it better and clarify with narrative writing instead of people's

01:26:31.500 voices. And so in every way, it refreshed me. And so now it's like something I really want to do.

01:26:37.320 And one of the, kind of share this, I'll attack on another story here, but this is one of the

01:26:40.620 greatest things I've ever seen in sports live. It has to do a little bit with this, having something else

01:26:44.360 going on in your identity, which was at the Vancouver Winter Olympics in 2010. Yeah, 2010.

01:26:51.660 I was up at the women's 1.6K cross-country sprint. They say sprint, but you go like way up a big hill

01:26:57.620 and come back down. And there's like four rounds. They're all in a day from the prelims to the medal

01:27:02.560 round. And this woman named Petra Majdic, Slovenian skier, had been one of the top ranked people in the

01:27:08.160 world for years and never medaled at world championships or Olympics. Totally snakebitten. Something would

01:27:13.080 always go wrong. They didn't bring the right skis. Oh, the technician didn't have the right wax,

01:27:17.340 ski brakes, like just freak stuff. And here she's favored to get silver, I think, in Vancouver. And

01:27:24.080 in the warmup, right before the first round, she slides off a curve, falls into a frozen creek bed

01:27:29.560 and bruises up all of her side. And they take her to do quick examination because she has to go to the 0.98

01:27:34.140 first round and say, okay, nothing's broken. It's just pain. Like if you're okay, you can go. And she goes to the

01:27:37.900 first round, qualifies on time when she was the favorite in that round, barely makes it through. 1.00

01:27:43.020 Second round, similar thing. They examine her. It's just pain. You know, if you can do it, you can go.

01:27:47.000 Third round, she falls down after the finish line, screaming. They have to carry her away and examine 0.99

01:27:52.040 her again and say, if anything was really wrong, we'd tell you you couldn't go. But if you can go,

01:27:55.940 you can. Final round, you hear her screaming every time she's like pulling the ground, gets the bronze 1.00

01:28:00.780 medal, just nips in there. Then they have time to examine her and find she broke a whole bunch of ribs 1.00

01:28:05.200 on her side. One of them broke off and punctured her lung. And so she came to the medal ceremony

01:28:09.980 with a tube sticking out of her chest in a wheelchair. And I remember she was, I talked to

01:28:15.540 her sports psychologist and they were saying, if we had known that she had broken all these ribs or

01:28:19.300 And had a freaking pneumothorax. Yeah. They weren't going to let her go to the medal ceremony. She was

01:28:23.660 like, I will die in the medal. I'm going, you know? And so they brought, wheeled her and they all said,

01:28:29.600 well, first of all, it's a kind of a testament to the mentality that if, if they told her it was an injury,

01:28:33.560 she wouldn't have done it. Right. But it's just pain. And I was talking to her sports psychologist

01:28:36.600 about what got her here. Like, what's the journey been like, especially after she's always having

01:28:40.360 this stuff going wrong. This must've been like, Oh, again. And he said, one of the most important

01:28:44.480 things was like diversifying her identity. She was getting so fixated on, and so much pressure.

01:28:49.580 He was like, you need to do something else. So he forced her to start building a house basically.

01:28:53.520 And that became like a task for her to do and a whole new thing and a new skill. And he felt like,

01:28:58.240 I thought maybe he was going to talk about some new type of cross training. He was like, no,

01:29:01.600 the building, the house took the pressure off. It gave her some other part of her identity. So

01:29:05.920 it wasn't just as the person who's like always having something go wrong. And I just thought

01:29:09.220 that was so interesting that his, what he felt was his main contribution was giving her a hobby

01:29:13.400 essentially. And that he thought that really allowed her not to feel like the pressure that

01:29:17.920 would break her in some way. You sort of alluded to it earlier with the story of the Nobel laureates as

01:29:22.480 well. I can't remember if it was, I think it was Francis Crick who said this, I could be wrong

01:29:26.420 that, and it might've been Watson, but I think it was Crick who said the key to doing great science

01:29:32.540 is always being a little bit unemployed. And I'm also probably paraphrasing that and bastardizing

01:29:37.620 it. It must've been said much more eloquently, but you get the point. The gist of it is great

01:29:42.700 insight in science comes from having time to wander in your mind. You said something a while ago about,

01:29:51.820 which made me very jealous by the way, and kind of pissed me off. You would read 10 papers a day

01:29:56.080 in the exploration phase of writing. And it's like, I'd give anything to read 10 papers a day today.

01:30:03.100 Instead, I feel fortunate at least to have a team that can read 10 papers a day for me.

01:30:07.260 I remember that bliss of, you just get to read the paper all day, every day, highlight it, take notes,

01:30:12.600 call the author, go to journal club, like, boo, like that was unbelievable. And we don't,

01:30:20.820 I mean, it's very difficult to be a scientist today because nobody's paying you to be just

01:30:27.640 thinking. Nobody's really rewarding you. Nobody's promoting you based on that. So we've created this

01:30:34.400 very perverse set of incentives. And in some ways, honestly, I consider it a miracle that there are

01:30:39.260 still really good science being done. Most science being done, by the way, just sucks. I mean, if we're 0.98

01:30:43.300 going to be brutally honest, if you pull up PubMed and you look at every one of those hundred thousand

01:30:48.920 papers every month that makes their way on the PubMed, I think 90 to 95% of them are absolutely

01:30:56.140 useless. Serve absolutely no purpose to our civilization. Do not advance natural knowledge.

01:31:02.700 Do nothing beneficial for us.

01:31:04.900 They increase the publication count of the person doing them, which is why they're doing them.

01:31:08.480 That's right. It's an economic tool of the journals and it's a promotional tool. So you don't

01:31:13.300 have any absolutely nefarious actors in the situation. You just have a system with such

01:31:18.120 perverse incentives that nothing good. Some of the predatory journals, but those are.

01:31:21.900 Yeah. Yeah. Yeah. Yeah. And I think a lot of the science is bad. I confess in range that I think my

01:31:26.800 own master's thesis was not good science and wouldn't replicate, but I didn't know, like I was rushed into

01:31:31.800 studying the specifics of Arctic plant physiology really quickly without learning what was actually

01:31:36.620 happening when I hit a button on the statistics program and how science has to be set up to make

01:31:41.240 conclusions. And I think that's kind of the norm for a lot of scientists. And I think there's 2005

01:31:46.020 paper by John Ioannidis that was titled why most published research is false that people kind of

01:31:50.200 wrote off. Now it looks like he's a genius, right? Oh, and now I hope I'm not saying something John

01:31:54.540 wouldn't want me to repeat, but I think John's been pretty vocal about this. Probably like seven or

01:31:58.660 eight years ago, I was having dinner with John and I said, John, I want to know bullshit answer on this.

01:32:03.320 What percentage of people who do science for a living, so they apply for grants, they get the money,

01:32:08.680 they do the thing. What percentage of those people shouldn't be doing that? Which doesn't

01:32:14.440 mean they can't work in a lab, but they can't be principal investigators. I mean, he said something

01:32:19.840 to the tune of 95. He's like 19 out of 20 people who basically for a living are in the business of

01:32:26.220 generating hypotheses, testing them experimentally and evaluating the results of those experiments.

01:32:31.520 That's the scientific method. 19 out of 20 people who do that can't do it and probably should not be

01:32:37.060 doing it. That's scary. But also I would guess that we don't want to try to narrow it down to

01:32:42.460 just the people that should be doing it because you'd have this purifying form of selection where

01:32:47.580 you have to like allow some failure. So this is kind of where I want to go with this thinking,

01:32:51.660 right? So why are we going down this tangent? So you've obviously written a lot about the importance

01:32:56.940 of even generalists within science. And you've given a couple of examples, which we should go into

01:33:01.600 about the hard problem that gets solved by somebody whose native scientific tongue is not

01:33:09.020 the one that is being addressed. So I want to talk about that, but I want to contrast it with this

01:33:15.380 other thing, which is I have this thesis that I harp on all the time, which is mostly my way of

01:33:21.680 communicating with myself to not get so upset at the lack of scientific insight in the world. And the

01:33:27.600 idea is genetics and evolution have never prepared us for science. So the scientific method is less

01:33:35.080 than 400 years old. Even the earliest signs of formal logic as a construct to describe the way we

01:33:41.680 think represents less than 0.1% of our genetic existence. So Atiyah, why would you ever get upset at

01:33:49.600 somebody who can't think logically? Why would you ever get upset when you pull up a story on Goop

01:33:55.840 that is written like the most insulting thing you've ever seen in your life with respect to

01:34:00.980 science? Why would you let that upset you, Peter? It's like you're expecting something that is

01:34:05.760 simply too much. And you might be assuming that the people writing for Goop are trying to get things

01:34:10.960 right. So that might be the assumption. That's part of the assumption. I would love to hear you hold

01:34:14.580 forth about Goop, to be honest. So I contrast this with there has to be some degree of rigidity and

01:34:21.780 formality in the training. I like to think of myself as a pretty good critic of science. I have

01:34:26.700 the ability to read papers and immediately, most of the time, figure out why these are total crap and 0.96

01:34:32.720 what all the biases are and all these other things. But to take any credit for that would be ridiculous. 0.98

01:34:37.860 I can't take any credit for that. That's simply because I was mentored by people and I went through

01:34:42.920 a formal type of training or informal type of training. Really, it wasn't formal and codified,

01:34:47.320 but it was informal. It's going to journal club. It's doing the experiments yourself,

01:34:52.300 putting something together, thinking you're slick and having people tell you,

01:34:55.700 hey, numbnuts, did you realize how many ways you screwed this up and how you've drawn the wrong

01:35:00.280 conclusion? So how do you balance that you have to go through that type of training with,

01:35:07.360 but sometimes you have to be an outsider? Yeah. And by the way, once you mentioned informal

01:35:11.560 training, I think we underestimate the importance of informal training and you have to kind of set up

01:35:16.220 cultures for it and things like that. So I have a master's degree in geological sciences and I got

01:35:21.620 a much better education in genetics writing a book than I did in a grad program. I wouldn't know

01:35:26.560 how to run like the equipment in the lab, but I basically kept like a statistician on retainer and

01:35:31.760 just to talk to me about a paper anytime. And that's an amazing way to learn. You're not just

01:35:36.060 being talked at, you have a specific question of why is this right? He's like, no, you can't use

01:35:39.220 this. And so it's an amazing way to have like informal learning. And you're right. I think the way to

01:35:44.340 balance that I think it's difficult because you're right, you do need some formal training in it.

01:35:48.160 Like in other words, the takeaway from your book is not that the person who's going to crack the

01:35:53.600 code on pancreatic cancer is currently working at an investment bank as a finance analyst who's

01:36:01.540 never taken a science course. That's not going to happen. Not even close. And I don't think you're

01:36:06.520 trying to suggest that, but I've heard people try to take your work and paraphrase it as, 0.98

01:36:11.160 oh, well, all these scientists working on cancer, you screw it, man. They're not going to figure

01:36:14.680 anything out. We need to, we need to go get the history majors to solve the cancer problem. And

01:36:18.760 I'm like, not a freaking chance without some modicum of scientific training.

01:36:23.800 Yeah. And I mean, I think the scientific community, I would say the tech community in particular

01:36:26.600 would do well to interact with historians. Like I think there are a lot of things they could learn.

01:36:30.200 Sometimes I feel like the Silicon Valley set maybe doesn't respect history quite enough or it'd be

01:36:34.280 useful for them, but that's a side point. My guess is that when people are saying that,

01:36:37.720 that just like go pick some person at random to do science, maybe they're taking that from

01:36:43.440 the chapter where I wrote partly about Innocentive, which was the VP of research and development at

01:36:49.660 Eli Lilly in the past. And this guy named Alf Bingham, who's that? I remember when I first

01:36:53.540 talked to him, he said, look, I'm an organic chemist. It doesn't have a carbon in it. I'm not

01:36:56.560 even supposed to play with it. Okay. I'm specialized. And I guess he realized at a certain point that

01:37:01.260 chemists at Lilly were getting so specialized that there were certain things they were great at,

01:37:04.340 but it also narrowed their view. So he talked a lot about this terminology in some of the business

01:37:08.820 literature, exploration versus exploitation. Exploration meaning essentially going and looking

01:37:13.040 for new ideas and solutions. Exploitation, once you find them, how do you make the most out of them?

01:37:17.240 Both of those incredibly important. And he said, basically the exploration phase is increasingly

01:37:22.080 found outside because people are so specialized. So they're just not covering as much ground.

01:37:26.680 And so he had this idea to just post online some of Lilly's problems in drug development. And at first

01:37:31.520 everyone was like, no way, proprietary information. He said, well, pick stuff. Well, nobody will know

01:37:35.260 what we're doing. And they said, who else is going to be able to solve this? So he posts them online

01:37:39.200 and like a third of them get solved. I remember one of his favorite memories was an attorney who solved

01:37:44.260 some important chemical synthesis project because he had worked on some tear gas copyright case or

01:37:51.000 something. And it reminded him of that. And so a third of those problems get solved, which is amazing.

01:37:56.420 Yeah. That story amazed me actually.

01:37:58.220 And in those problems, it does tend to be, because they've selected for problems that have

01:38:02.280 stumped the specialist, right? So it does tend to be the more likely the problem is to get solved,

01:38:05.960 the diversity.

01:38:06.840 And I think my question on that to really double click on it is, did the generalist actually solve

01:38:11.620 the problem or did the generalist just come up with a clue that completely changed how the specialist

01:38:16.720 went about approaching it? Like in the case of that example, the attorney doesn't actually know how to

01:38:22.420 completely synthesize the molecule. What they're basically saying is, you guys are looking at it this way,

01:38:26.800 stop. Rotate 90 degrees over there and turn around. And I think the answer is over here,

01:38:32.880 to which case either a new set of specialists could do it. Is it sort of a bit of that hybrid?

01:38:37.720 Yeah. And actually, Innocent have evolved to where they give different... Now they help other

01:38:42.060 operations post their questions in a way that'll attract what they call solvers. And

01:38:46.420 there are different monetary rewards depending on what kind of contribution it is the person makes.

01:38:51.800 And sometimes he did get people sent in powders and stuff. They synthesize stuff on their own. Other

01:38:56.240 times it was much more like you're taking the wrong approach. And here's another thing to think about.

01:38:59.660 But they were still chemists that had to do this, right?

01:39:02.140 Not all of them. Not all of them. But there was a lot of that. Sometimes they'd be chemists from

01:39:05.800 some other area, but sometimes it was totally random people. But again, you're farming it out to the

01:39:09.880 whole world. People have... Not the whole world. I mean, most people don't know about Innocentive,

01:39:13.640 but a huge number of people. And I think he was surprised sometimes things would come from

01:39:18.380 people who worked with machinery but weren't really formally trained in it, but had a lot

01:39:24.440 of experience. And so I think sometimes there were true outsiders. But most of the problems don't get

01:39:30.080 to the Innocentive stage anyway. They're being handled by the specialists. And I wouldn't extrapolate

01:39:34.420 Innocentive to mean that specialists aren't incredibly important. I mean, my broad view of this is the

01:39:39.140 same as physicist Freeman Dyson where he gave this great speech where he said, we need birds and

01:39:43.160 frogs. The frogs are down in the mud looking at the granular details. The birds are up above. They

01:39:47.160 don't see the detail, but they're integrating the knowledge of the frogs. And he said, the world is

01:39:51.720 wide and deep. It'd be stupid to say one's better than the other. You need both for a healthy ecosystem. 0.98

01:39:56.140 His concern was we're telling everybody to be frogs. And so we're not having birds. And that's kind of

01:40:00.280 how I conceive of it. And I think the biggest impact would come from what Arturo Casadoval, who's this kind of

01:40:07.640 one of the prime characters of the last chapter, is trying to do, where adding range to people who

01:40:12.340 are within science already. Like these people are being formally trained. He's just backing up the

01:40:16.800 formal training instead of jumping right into kind of the reductionist, studying the body as a machine

01:40:22.460 sort of thing. It's like the initial classes in the program he's pioneering. And he's like one of the,

01:40:26.900 I think his H index, which is a measure of his productivity as a researcher, surpassed Einstein's

01:40:30.980 recently, which isn't fair because people publish a lot more now, as you alluded to for their careers,

01:40:34.780 but it still puts him in very rare company. And so he's starting people with how do we know what is true

01:40:40.100 and the anatomy of scientific errors and also how errors have sometimes led to breakthroughs. And so he's just

01:40:45.740 like backing up the training into these broader concepts and saying, you can learn the more specific

01:40:50.560 didactic stuff later, but if you don't learn the broader conceptual stuff, you never get it because

01:40:56.360 you're only going to get more and more and more specialized from here. So I think one way we can approach

01:40:59.960 this is just by backing up the training basically. Cause I don't think we have to worry about

01:41:04.360 people will learn the specialized stuff in their field just by doing it and being there,

01:41:08.420 but the other stuff, but that's what we kind of teach them and start them with. But the other

01:41:11.980 stuff like scientific thinking, how does it work? What constitutes evidence? They'll never get that

01:41:16.440 stuff if they don't get it early in training. Yeah. I sort of cite the examples you've given when I

01:41:20.980 get asked a question a lot, which is someone going to college who wants to go to medical school

01:41:25.900 saying, what is the best thing to study in college to which I don't know the answer, but my advice is

01:41:30.480 anything but pre-med. So you couldn't really do anything worse. And I'm sorry, because I know

01:41:34.380 there's somebody listening to this. I'm sure who's doing pre-med, who's going to go into medical

01:41:38.060 school. And my only take for you is, okay, I'm sorry that you're in pre-med right now, but make sure

01:41:43.580 you spend a lot of time doing non pre-med stuff in college. You still have the bandwidth to take other

01:41:48.480 classes. You should be doing so liberally, but you are better off being a history major who goes to

01:41:53.520 medical school than a biochem major who goes to medical school. In my opinion, I could be wrong,

01:41:58.260 but having seen enough people go through it, you're going to learn as much biochemistry as you're

01:42:02.140 going to need to learn when you get there, both interpersonally and frankly, in the breadth of

01:42:06.600 thinking, you'll be better off if you studied something else. I mean, a lot of Arturo's argument

01:42:10.260 too is I saw him make this argument talking about he went from Einstein to Johns Hopkins School of

01:42:15.780 Public Health because they're allowing him to start this new grad program. And I saw him on a panel

01:42:20.520 about the replication crisis in science, this problem with a lot of work basically not being

01:42:24.340 true. And the head editor of the New England Journal of Medicine said, you can't do that.

01:42:28.580 Training is already too long. And Arturo said, it's clearly not working. He said, I'm saying drop the

01:42:33.340 didactic stuff because it's not been working. I think he actually pointed out that the New England

01:42:36.860 Journal of Medicine had the highest retraction rate in some study also. But so I thought that was

01:42:41.360 interesting where he said a lot of that stuff is in one ear and out the other anyway, if it's really

01:42:45.380 didactic, people don't even know if they're going to need it or when they're going to need it.

01:42:48.420 And so I thought that was an interesting take that he said, you can just drop some of that other

01:42:52.060 stuff. But to your point about telling people not to do pre-med, when I was at Sports Illustrated,

01:42:57.300 not so much anymore, but when I was there, I would get asked by a young aspiring journalist,

01:43:01.780 what should I do if I want to work as a sports writer? Should I major in English or journalism?

01:43:06.240 And my first instinct was to say journalism. If you know what you want to do, get a head start.

01:43:10.900 Second instinct, English. And then if I thought about it, I'd be like, well, I majored in geology and

01:43:14.420 astronomy, so I have no idea what to tell you. But stats course, biology course,

01:43:18.200 never hurt anyone. You should do one of those because you'll learn the job by doing the job.

01:43:22.740 And that's the only way you'll learn the job. That is the challenge of giving advice, isn't it?

01:43:26.200 Because I don't feel this way at all anymore. I'm totally comfortable with my nonlinear path to

01:43:31.200 doing what I do. But I spent a great deal of time frustrated that I couldn't figure out sooner in

01:43:38.100 life what I wanted to be when I grew up. And I thought of all the time squandered. But again,

01:43:43.020 that was through the paradigm of, you only had one to two 10,000 hour windows. You didn't take

01:43:49.660 them. You spent them doing something you will never do again. I had my 10,000 hour shot and I

01:43:53.880 put it in the wrong thing. And now I never do it again. But I think that speaks to probably the

01:43:58.940 insights of your book, which is you're discounting a bunch of things you learned in doing that.

01:44:03.840 And also, sorry, I didn't mean to interrupt you. Were you going to? No, no, no. Yeah. And also I

01:44:07.000 think your insight into yourself, your skills and interests in the world are constrained by your

01:44:11.920 roster of previous experiences. As Herminia Ibarra, who studies how people find what's called match

01:44:16.620 quality, the degree of fit between their interests and abilities in the work that they do, which turns

01:44:20.440 out to be really important for how likely they are to burn out, for their performance, for all these

01:44:25.160 sorts of things. And we really underestimate it. She studies basically how people seek this out in

01:44:30.320 transition careers. And she said, we learn who we are in practice, not in theory, which means

01:44:35.820 there's this kind of cultural idea and lots of career gurus and personality quizzes that kind of

01:44:41.960 seek to convince you that you can just introspect and know what you should do putting in that time.

01:44:46.020 But in fact, the way we learn about what we want to do is we have to do stuff, act and then think.

01:44:50.680 You have to do stuff and then reflect on it. And that's the only way you figure it out. And match

01:44:55.000 quality seems to be so important that spending some time in that experimentation is worth it.

01:45:00.140 So one of the, that I think is sort of representative of that, this economist who found a natural

01:45:04.980 experiment in the higher ed systems of England and Scotland, where in the period he studied,

01:45:08.980 students in England had to pick a specialty in their mid-teen years to apply for a certain

01:45:12.900 program in college. The Scottish students could keep sampling throughout university. And he said,

01:45:17.000 who wins the trade-off? Otherwise, the systems were very similar. Who wins the earlier late

01:45:20.700 specializers? And what he found was the early specializers do in fact jump out to an income lead

01:45:25.040 because they have more domain-specific skills. But the late specializers sample more things.

01:45:30.020 And when they do pick, they have better match quality. And so they have higher growth rates.

01:45:34.220 So by year six out, they fly by the early specializers. Meanwhile, the early specializers-

01:45:38.860 That's based on the economics.

01:45:40.080 Yeah.

01:45:40.340 What about when you look at things that probably matter even more, such as happiness and

01:45:43.540 satisfaction?

01:45:44.160 He was only looking at finance and career switching. So, because he was looking at huge numbers of

01:45:47.940 people. But we can talk about fulfillment in a sec. So the early specializers then start

01:45:51.640 leaving their career tracks in much higher numbers, even though they have much more

01:45:54.160 disincentive from doing so. They're made to pick so early that they more often make a wrong choice.

01:45:58.640 So the return to match quality was higher than the return to getting a head start in domain-specific

01:46:03.940 skills. The Dark Horse project, which I wanted to get around to anyway, has to do with fulfillment

01:46:08.460 because I think you should tell more of your story because it sounded really interesting and

01:46:12.120 I want to hear more of it. But that project, the dependent variable was fulfillment and it

01:46:16.160 was about how people go on.

01:46:17.840 Yeah. Well, before we started the podcast, I said there's this thing I've always tried to

01:46:21.300 explain when I've told my story to high school kids or something like that. It doesn't really

01:46:26.380 make sense, which is growing up, all of my energy went into boxing and martial arts and

01:46:31.320 I didn't do anything in school. I was super mediocre. And then I had this awesome teacher

01:46:37.500 who, when I was in grade 12, in Canada, you say grade 12, not 12th grade. Sounds stupid. 1.00

01:46:44.080 So weird. I can't believe we can even get along with you guys. 1.00

01:46:45.780 It's hard that we can even have a discussion. And he called me in one morning and he said,

01:46:49.040 hey, I heard you're not applying to university and stuff. And I said, that's right.

01:46:52.220 And he certainly didn't bust my chops. In fact, I wrote about him after he died and how amazing

01:46:57.260 it was that he just knew what to say. He knew the right thing to say at the right time, which was,

01:47:01.340 hey, I totally get it. When I was your age, all I wanted to do is play in the NHL. And it was the

01:47:04.900 only thing that mattered to me. And you really ought to, don't let anybody tell you that your

01:47:08.780 dream of being middleweight champion in the world is a dumb idea. It's not. But he said, but it was 0.51

01:47:13.200 almost like a Steve Jobs moment. Like one more thing. I think you have a gift for mathematics and

01:47:18.740 you should at least entertain the idea that maybe that's your calling is more in terms of math than

01:47:24.660 it is in fighting. And I remember that day as clear as, I mean, that's so long ago. It's more

01:47:29.900 than half my life ago was that moment, but it did change the course of my life completely. And then I

01:47:35.140 did come back to high school to finish and did better than anybody expected I could ever do. And

01:47:40.240 I've always assumed the reason I was able to do that is I simply took the work ethic of exercising

01:47:46.800 six hours a day, which was sort of what I did in high school. I was, I would run five to 13 miles

01:47:51.720 every morning, 400 pushups before bed every night. And everything in between was training around that.

01:47:56.620 And I just applied that to calculus, algebra, physics, geometry, et cetera. So that sort of gets

01:48:01.720 into the Duckworth grit stuff. The question I guess I have in that is, had I taken the grit of age 13 to 18

01:48:09.540 and basically with no other cognitive capacity turned it into then doing well in school?

01:48:15.480 I definitely wouldn't say no other cognitive capacity because, well, I think those things

01:48:20.080 take a different certain cognitive capacity, that kind of discipline. But clearly the teacher

01:48:23.900 recognized something in you that you did not recognize in yourself. It didn't just pick you

01:48:27.640 out randomly. And I assumed. Well, especially because I wasn't even the top student in mathematics

01:48:31.380 at the time. I wasn't even near the top. So yeah. So obviously you saw something that you didn't

01:48:36.200 yourself see. And so it seems to me like you had this training, whether that was something that

01:48:40.400 was part of who you are naturally or something that motivated you that you could take that and

01:48:44.700 transfer that to then something that had better match quality for you. And so it's a combination

01:48:48.420 of you using certain skills and approaches that you had learned and somebody helping you find

01:48:52.500 better match quality, basically. I think that's like an explosive combination in a good way.

01:48:57.840 But it is interesting. I wonder what he saw because he was right. But did you have any inkling that that

01:49:03.260 was the case at the time? Or was that total news to you that he thought you might have math?

01:49:06.680 No. In fact, I found math incredibly frustrating because it wasn't something you could BS your way

01:49:11.880 through. At least in English class, I remember this well because I'm actually still very close

01:49:16.140 to my English teacher from high school. I always managed to find a way to weasel through by writing

01:49:21.060 every essay I wrote was on Muhammad Ali or Mike Tyson or Marvin Hagler or Jimi Hendrix. I basically

01:49:27.260 screwed my way through English by always figuring out a way to read a book that I was interested in

01:49:32.500 and figuring out a way to read. But you couldn't do that in math. You sort of had to do it. So

01:49:36.460 what I remember is that math frustrated me more than anything. But I wish Woody was still alive

01:49:41.540 for many reasons. But one of them being, I wish I could ask him why he said that.

01:49:45.820 Yeah. No, it's curious. I mean, and I think a lot of people find math more frustrating in English

01:49:49.760 class. It's like English class is more subjective and you can kind of get through even if you don't,

01:49:54.160 a lot of people, even if you don't know what you're doing. But that is, that is interesting. I wonder

01:49:57.380 what he saw. And I think that's the reason I said, I think it resonates with the Dark Horse project,

01:50:01.920 which was this project by these two Harvard researchers to sort of figure out how people

01:50:06.360 who find fulfillment in their work go about it in a wide range of careers. Essentially,

01:50:10.260 there was a huge variety. But most of the people did not stick on their first sort of dream,

01:50:15.700 basically. There were some, but it was a small minority. Most of them had this, they actually

01:50:19.880 called it the Dark Horse project because people would come in and say, well, don't tell anybody to

01:50:23.580 do what I did because I started this other thing and then I, something else came up or something

01:50:28.360 random happened or I realized there wasn't what I want to do. So I had to start my own thing.

01:50:32.360 So they also felt sort of sheepish about their non-linear path to get there.

01:50:36.740 Yeah. And they saw themselves as having gotten lucky and come out of nowhere, which you do have

01:50:39.320 to get lucky. Luck is important, but they all saw themselves as having come out of nowhere,

01:50:43.020 which is why they called it the Dark Horse project. What they kind of had in common was this,

01:50:46.980 they would sort of respond instead of sticking to like an ironclad long-term plan,

01:50:51.440 they would respond to their lived experience by zigging and zagging and kind of finding what they

01:50:54.740 were good at, whether that was someone helping them figure that out, which I think is often the case for a

01:50:59.940 younger person or them being older and maturing and sort of realizing that. And it sounds like that

01:51:04.240 kind of happened to you in some way. You view yourself as coming out of nowhere because you had

01:51:08.580 a talent and also some transferable skills that could come together in a way that you hadn't

01:51:12.920 conceived before. And it worked really well. So I think you'd be kind of the type for the Dark Horse

01:51:17.900 project.

01:51:18.260 I mean, to take something you said a moment ago and couple it to this. So let's say we fast forward

01:51:24.180 whatever 18 years and your son says, dad, I don't want to go to college. I'm going to learn a bunch

01:51:28.080 of stuff that A, I don't need to learn because I could teach it on my own. I'm going to get a degree

01:51:33.280 in something that doesn't necessarily imply what I'm going to do with myself thereafter. The letters

01:51:40.460 don't really matter after my name. And by the way, notwithstanding the fact that maybe by the time

01:51:45.480 our kids are going to college, the debt that they'd incur to do so is itself more debilitating

01:51:51.600 than anything else. So can you make a case that one does not need formal education at that level

01:51:59.200 to go far? I mean, notwithstanding the fact that sometimes you need it from a professional

01:52:02.420 standpoint, like you can't become a lawyer without going to school. I don't see that changing and

01:52:05.880 things like that.

01:52:06.980 At least in California, you just have to pass the bar.

01:52:08.740 That's a great point. You can't become a doctor, I guess, without going to school. And

01:52:12.500 so does this change the way you think about higher education?

01:52:16.200 Yeah. And gosh, I don't even know where to start with this question. I hope that college

01:52:19.540 looks different 18 years from now than it does now, because if we keep ramping up people's debt,

01:52:23.220 then we will make sure that they fall prey to the sunk cost fallacy, where the more you invest in

01:52:27.840 something, one of my favorite writers, Maria Konnikova, wrote a book about con men.

01:52:31.940 And one of their strategies is they start with asking for small things.

01:52:35.960 Oh, I thought you meant con men.

01:52:37.500 Oh, no, no, not con men. No, no. Con men, playing confidence games. Yeah,

01:52:42.180 the book's called The Confidence Game. And she's a psychology PhD and a great writer. And

01:52:47.160 she notes that they'll start with these small, lots of small asks, because the more you invest,

01:52:52.360 the more likely you are to fall prey to the sunk cost fallacy of then saying like, well,

01:52:55.180 I've already put some in, so I should keep going. Even when to an outsider, it's like clear that it's

01:52:58.580 a disaster. And I think the more debt we saddle people with, like that study we were just talking

01:53:03.680 about with the higher ed systems in England and Scotland, the English students who specialized

01:53:06.800 earlier have more disincentive from quitting, even though they should. When they do quit,

01:53:10.720 their growth rates are then higher because they're quitting in response to information

01:53:13.940 about themselves that they've learned. And I think the more time and debt we saddle people

01:53:19.280 with, the more we make sure they will not respond to match quality information. And instead,

01:53:23.700 we'll say-

01:53:24.600 Yeah, you blunt their receptivity to the signal.

01:53:26.620 That's right. And I think we want to, Steve Levitt, the economics economist, did this interesting

01:53:30.640 study where he had people flip a coin to determine major life decisions. And the most commonly asked

01:53:35.080 question was, should I quit my job? And what he found was the people who flipped, I think it was

01:53:39.220 heads and changed their job were better off when he checked in with them later. And so I think you

01:53:43.580 want like as little friction as possible to people job changing. In fact, when I was watching one of

01:53:47.920 the Democratic primary debates recently, and people were talking about universal healthcare,

01:53:52.380 I would think one of the advantages might be that it would lower friction to job changing because

01:53:56.080 you're not as worried about, so that maybe people can shuffle around more and have better match

01:53:59.680 quality. So I think something has to be done about the debt situation. I think there's plenty of

01:54:04.740 evidence that for a lot of students- So the economist Brian Kaplan wrote a book called

01:54:08.640 The Case Against Education. I certainly don't agree with- And it's higher education specifically.

01:54:12.540 I certainly don't agree with everything in the book, but I think it's provocative and a rigorous

01:54:16.060 take. And his argument- So we know that some part of college education is signaling to the job world

01:54:22.060 that is not about anything that you've learned. It's just about, I am smart enough and-

01:54:26.320 I'm serious enough.

01:54:27.340 It's like the minor leagues for the job world. So we're doing them a favor because they don't

01:54:30.440 have to be good scouts. And basically what he was saying is nobody says that either signaling

01:54:37.660 or learning is all of the effect of college. But his argument was that he thinks signaling is like

01:54:41.520 80% of it, whereas other people would say, oh, maybe it's 20%. So he thinks most of it is just

01:54:45.820 you need this credential to signal to the work world that I'm okay. And I don't know if he's right

01:54:51.160 about how much it is, but I think it's probably more than people intuit because research seems to

01:54:56.260 suggest that it doesn't change people as much as we might think in some ways. What I do think is

01:55:00.960 important about it though, is that it does give that kind of sampling ability. Like I didn't know

01:55:04.840 about it. So the Scottish students in that study also much more often the late specializers end up

01:55:09.940 studying something that wasn't offered in their high school because they didn't know about it. I

01:55:13.400 didn't know about the stuff I studied until I got to college. And so without it, I would want there

01:55:19.140 to be some other mechanism for sampling. But I think that kind of communication technology that we have

01:55:25.240 now and the internet may, if we use it smartly, it can expose you to a lot more stuff. And I think

01:55:30.720 there's huge potential of online courses. I love some of these online courses. So I do think while I

01:55:37.620 would want him to be able to have a sampling period, my kid, I think I don't know that the same kind of

01:55:42.500 college model that we have now will be the answer. And I think there are other things that could expose

01:55:46.480 him to more other things. But if society forces the signaling to continue, then what can you do?

01:55:51.540 Then it's just like, you have to pay your dues because you have to pay your dues. The way I

01:55:54.920 see my role as a parent, we mentioned Angela Duckworth, by the way, I should mention the

01:55:58.760 week that my book came out, I subscribed to her newsletter. Her newsletter was titled Summer is

01:56:02.660 for Sampling. And she said, kids shouldn't be too gritty until they figure out what to be gritty in.

01:56:07.520 And she says, it took me a decade of trying stuff to figure out. So it really speaks to a broader

01:56:12.420 point here, which is a lot of these insights that we sort of latch onto are slightly out of context.

01:56:19.020 Yeah, definitely. Definitely. And I mean, so now if her point is, you should be gritty when you

01:56:24.020 should be gritty. I'm totally on board with that. I view my roles, I write a little bit in that section

01:56:29.060 about grit, about something I should have written more about, about the way the army adjusted when

01:56:33.740 they realized in an industrial economy, and they had this strict up or out structure, it was great

01:56:37.460 because you did want to be specialized because work next year did look like work last year. And there

01:56:42.700 were huge barriers to lateral mobility, but knowledge economy comes along. And now there's tons of

01:56:46.700 lateral mobility for people who can engage in knowledge creation and creative problem solving.

01:56:51.100 So their highest potential officers started quitting if they didn't have any agency over

01:56:56.040 their career matching. And so first they tried to throw money at people to retain them and people

01:57:00.440 were going to stay, took it. People were going to leave left anyway, half a billion dollars of

01:57:03.940 taxpayer money down the drain. Then they started programs like one they call talent-based branching,

01:57:08.580 where instead of saying, here's your career track, go up or out, they say, we're going to pair you with

01:57:12.160 a coach and try this one career track and reflect on how it fits you. And then try these other two

01:57:16.940 and these other two and we'll triangulate a better fit for you. And that improved retention more than

01:57:20.480 throwing money at people. They've basically built in a match quality sampling system. And so maybe we

01:57:25.460 could do stuff. That's how, frankly, how I view right now my parenting role as being the coach in

01:57:29.740 the math talent-based branching system for my kid. Facilitate, help him know that a lot of things are

01:57:34.160 available and get the maximum amount of signal about himself from each one. And I think there are ways

01:57:38.680 that that can be done possibly more effectively and a lot more cheaply than the formal education

01:57:45.700 structure. How do you know how long to push? So give you two examples with my daughter. And I feel

01:57:50.760 so bad sometimes telling these stories about her because one day she'll listen to this and she'll

01:57:54.300 be like, dad, I can't believe you embarrassed me. But when she was five, she said, I want to play the

01:57:58.920 drums. And we've had lots of parents who have interacted with us who said, are you crazy? How did you 0.93

01:58:03.880 listen to her? But we were like, oh, we just did. So we got her like a set of cheapo little toy drums 0.98

01:58:07.640 just to see if would she actually do anything. And she did. She wouldn't stop wailing on them. 0.99

01:58:11.620 But no, she was four when she said that. Then when she was five, she said, no, I really want to keep

01:58:14.940 playing. So I was like, okay, fine. So we got a drum teacher, got her a real set of drums and away

01:58:20.520 she went. And here she is now. She's 11. She still drums. She loves it. She's, I mean, she's really

01:58:26.520 good. I think I can say that without too much bias, just objectively based on what her teacher tells

01:58:30.640 me, which is like, she drums better than most anyone. And she's only 11 and we never pushed her.

01:58:35.680 Right. So it was just, she just wanted to do this. There was one time when she kind of wanted to quit

01:58:41.060 when she was about seven. And we talked to her teacher and said, Hey, let's spend way more time

01:58:48.860 just letting her play Taylor Swift songs than doing scales for a while and see what happens. And then

01:58:54.600 that was great. Like it all came back into the mix year ago. She starts taking tennis lessons.

01:58:59.000 I think her tennis teacher is the most awesome guy. He's like this young Russian kid. He's so fun. He's

01:59:05.240 so smart. I've, I've watched them. I mean, he's just has a beautiful way of explaining things to

01:59:11.000 her. She never feels bad. Like he never hammers her, but he's strict. If she's screwing around,

01:59:17.620 he tells her, she's like, I don't want to play tennis anymore. And I'm like, Olivia, are you

01:59:22.500 freaking crazy? Do you know what I'd give to have played tennis when I was a kid? And I don't get into 0.95

01:59:27.100 the sob story of like, we didn't have the opportunity for private tennis lessons or whatever,

01:59:30.580 but it's more the tennis is such a beautiful sport. You will be able to play this forever.

01:59:36.340 How can you not want this? And I'm really torn. Do I push her to continue taking lessons? And I'm

01:59:43.180 not saying like you have to go and even compete. It's just sit here and play tennis for two hours

01:59:47.840 a week. You and mom play. It's fun. Why wouldn't you want to do this? Or do I just say, well, she only

01:59:54.480 wants to play basketball, which is the only sport she wants to play right now. Should I just, I don't

01:59:57.640 know if this is a great example, but I think you get the point, right? It's like, there's a part of

02:00:00.420 me that thinks it's really good for her to keep playing tennis. Yeah. Yeah. And she's crazy to 1.00

02:00:04.480 stop. Obviously you're saying she'll regret it. It sounds like it. That's my potentially stupid

02:00:08.940 fear, which is you're going to be 51 day and you're not going to be playing basketball because 0.99

02:00:13.360 nobody's playing. No 50 year old woman is running around playing pickup basketball, but you will still 1.00

02:00:18.680 play tennis. So why wouldn't you continue to learn this now? Well, you have this critical window in

02:00:23.660 which you could get good at this sport. Yeah. And she could play tennis at startup at 50 1.00

02:00:27.620 just wouldn't be as good at it. Exactly. Yeah. And you're oriented toward achievement. So she might

02:00:31.860 be a great drummer and a crap tennis player. She starts at 50. She might still like tennis though. 0.97

02:00:35.660 Yes. And I'm trying to balance the, yes, I think that's the part of it. I can't let go of this idea

02:00:40.420 that I'm not great at anything. It would be really cool if my kids had the potential to be great at

02:00:46.760 something world-class. Like, wouldn't it be cool if she could be a world-class drummer one day? I have no

02:00:51.820 delusion. She won't just the interest or even the talent to be a world-class tennis player, 0.78

02:00:56.480 but I still think, why wouldn't you? You have this wind, I think I'm still stuck to this idea

02:01:01.060 that there's a window in which you can assimilate skill, be it language, music, sports, that to go

02:01:08.280 a little further. If she said, I don't know, anyway, that's sort of. And when she's younger,

02:01:12.280 also you have more time to put into it and you're a lot less concerned about making mistakes. So you'll

02:01:15.940 throw yourself in and practice in a way that you probably won't when you're older in most things.

02:01:20.000 I mean, that's a great question. It sounds like in drums, you found a way to sort of,

02:01:23.860 when she was saying, I don't really want to do it anymore to keep her foot in by sort of changing

02:01:27.060 what she was doing. Right. And it was just this one little audible called on the line of scrimmage,

02:01:30.820 everything was fine. But also it would have never even occurred to me to put a kid in drums. Like

02:01:35.200 this was 100% her insisting on it. And there was no resistance from us. I was like, great. But if

02:01:40.860 she'd said piano, violin, fill in the blank, we would have been equally interested in letting her pursue

02:01:46.040 that. So maybe it just comes back to this broader issue of, is there an age at which parents

02:01:51.520 should force some of the sampling? Is there an age at which parents say enough is enough? You've

02:01:57.500 sampled enough. If you don't want to play tennis anymore, don't play tennis. I should make clear,

02:02:01.320 I don't want to prescribe diversification any more than I want to prescribe specialization. Right. And

02:02:06.680 this gets to something I mentioned before, which is I think we tell the tiger and the Mozart stories a

02:02:10.520 little wrong. Yeah. Let's do that story. Okay. So tiger said in 2000, his father never asked him to play

02:02:16.120 golf. So his father did put a putter in his hand when he was seven months old, but he wasn't attempting to make

02:02:19.680 him a golfer. He was just giving it to him as a toy. And he responded. So he said in 2000, my father

02:02:25.500 never once asked me to play golf. It was always me asking him. It's the child's desire to play,

02:02:30.860 not the parent's desire to have the child play that matters. That resonates with the work of this woman,

02:02:35.080 Ellen Winner, who's maybe the world's authority on, on prodigies of that nature, that they're usually 0.98

02:02:41.000 driving their parents crazy, not the reverse. And that there's not really. Yeah. I mean, Wayne Gretzky is a

02:02:46.680 great example of that. You listen to these interviews with his parents, like they couldn't

02:02:49.940 get this kid to come in for dinner. And what Ellen Winner, I think she alludes to some research. 1.00

02:02:54.560 I don't think this research was exactly hers, but it's really a problem when someone's parents,

02:02:59.560 if they're low socioeconomic status and they have a prodigy and the person has this incredible drive,

02:03:04.300 a master, whatever it is they're doing. And the parent can't accommodate that because it really is

02:03:07.960 hard for the parents. Obviously tiger's father responded to his very unusual display of interest

02:03:13.420 and prowess. Mozart, probably the second most famous example. Tell people just what the,

02:03:18.180 what's the mythology around Mozart that, Oh, that his father basically was like Tiger

02:03:22.620 Wood's father, except for music that he started a very young age. Shoved a violin into the crib and.

02:03:26.680 Yeah, exactly. And so I was going through these letters and one that I remember really well

02:03:31.760 about Mozart's early life was a musician who would come over to play with Mozart's father,

02:03:35.300 who was a musician. And he recalls young Mozart coming in with the group of adults and saying,

02:03:42.380 I want to play the second violin part. And Mozart's father says, go away. You haven't

02:03:47.600 any lessons. You obviously can't play violin. And then the letter writer, his name was Andreas

02:03:53.140 something. I can't remember the last name says little Wolfgang started crying. And so I said,

02:03:58.320 okay, I'll go play with him in the other room. And his father says, but play quietly. Don't disturb

02:04:01.900 us. Next thing you know, they hear the second violin part coming from the other room. And so they

02:04:05.440 walk in and look and young Wolfgang Mozart's playing the second violin part with made up fingering

02:04:11.100 because nobody's taught him the fingering. And then the letter, this part I remember verbatim,

02:04:14.660 he says, young Wolfgang was emboldened by our applause to insist that he could also play the

02:04:19.100 first violin part. And so then he goes and does that. And that's when his father says, holy crap.

02:04:23.580 Yeah. And responds to it. So neither of those cases were they like manufactured the way we tell 0.99

02:04:28.180 that story. So I don't think we should be scared about missing a tiger or Mozart because the best way

02:04:32.480 to find one of those people is probably still to expose them to a bunch of things and see if something

02:04:36.580 lights their fire that way. I mean, I think of Mozart. That's a great point. That's a great point.

02:04:40.780 Yeah. I think those, I'm sure there's a lot of those people who never got exposed to the thing

02:04:44.260 that would have ignited what Ellen Winner calls that rage to master. And most people, you're not

02:04:49.080 gonna have a lot of tigers and Mozarts no matter what. And obviously those are exceptions. But I think

02:04:52.280 that's important to keep in mind. To go back to when do you allow your kid to quit? This is something

02:04:57.040 I've seen Angela Duckworth has tried to respond to a number of times. It's very difficult. She'll say things

02:05:01.160 like, don't quit on a bad day, which I think is sensible. But also, should you quit on a good day?

02:05:07.160 It's kind of hard to know what is the calculus for when you should quit. Like if you're having tons

02:05:11.400 of bad days, maybe you should quit on a bad day. And Levitt and Dubner have written about this,

02:05:15.980 right? I think it's in one of their, maybe their third book. Well, Levitt is always saying my best

02:05:20.000 thing is I know when to quit. Like everything, like fields of study, projects, whatever. Yeah. By the way,

02:05:24.960 as a funny aside, you know, Steve has tried very hard to get me to take up golf.

02:05:28.660 No, really? Yeah. So he's a very good golfer. A lot of people don't know how good a golfer Steve

02:05:33.220 is. Incredible. And he's convinced that taking sort of a freakonomics approach to golf, you could

02:05:39.640 take a novice and make them really good in a short period of time. We've had very serious discussions

02:05:44.980 about if I would give him one year, he could turn me into a really good golfer as someone who's never

02:05:50.580 even touched a golf club. I don't know if that's true, but I'm curious enough that I'd love to try it

02:05:55.480 if I had the time. Because when he's explained to me the logic behind it, like how many steps you can

02:06:01.920 shortcut if you're purely optimizing around this thing and how he's figured out a bunch of these

02:06:07.240 hacks, I'm like, oh, it's kind of fun. It's not going to happen with me, I don't think. But 0.99

02:06:10.740 yeah, probably don't have time to do that. But if you did take him up on that, like that guy we

02:06:13.760 talked about earlier, Dan McLaughlin, who dropped his job to try to become a pro.

02:06:17.180 Right. Who got 7,000 hours into it.

02:06:19.060 I think he got better. So he was getting way better at first, like anything, the learning

02:06:24.080 curve is. But I think how quickly he got better, and he had never golfed before ever, was surprising

02:06:29.500 to most people, how quickly he got better early on. And so I think doing that plus, I mean, he was

02:06:35.120 practicing, that was like his full-time job for the time. But if you did that for a year, plus

02:06:39.780 whatever hacks he's talking about, I bet you'd pass most of the amateur golfers out there. Because

02:06:44.160 most of them are just going and like swatting and it's like stress relief. They're not even really

02:06:46.880 trying to get better a lot of times. But I bet you'd surprise yourself with how good you'd get,

02:06:51.820 but I'm not a huge golf fan. I'm putting all of that energy into the things that I'm currently

02:06:56.540 doing. But this idea of quitting, it is really an underappreciated thing, isn't it? Because

02:07:01.320 sometimes it's time. It really comes down to time and exposure, and that's the most valuable

02:07:06.000 resource we have. But the when to do it is still the million-dollar question.

02:07:11.000 Totally. I mean, I think that's why I like this model of talent-based branching that the Army

02:07:14.800 used so much because they pair someone with a coach to help them make that decision. And it's still

02:07:18.340 not a science. And I also like they call it talent-based branching because basically it's

02:07:22.620 coached quitting, but they use a different term so it doesn't sound bad. But I don't know what's

02:07:27.700 perfect. I mean, I think someone, a mother asked me recently, said her son is really good at violin

02:07:32.380 and now he just wants to quit. And I think he just wants to play tennis. And she was saying,

02:07:36.580 but he's so good and I know he'll love it when he's older. And the best thing I could think of was

02:07:40.940 try to help him keep a foot in. Whether that's playing Taylor Swift songs or whatever, maybe

02:07:46.080 he just needs to back off and try something else for a while. Can you keep a foot in that pool so

02:07:50.540 that he doesn't totally detach from it and when he feels regenerated or whatever, can go back to it.

02:07:55.540 We know that progress comes from alternating stress and rest and things like that. So maybe

02:08:00.120 they just need a break or maybe they need to try some other type of training. I think that's

02:08:03.040 undervalued. If someone wants to quit and say, oh, either you have to stick with it or quit.

02:08:07.400 But the approach you took was, let's try varying what you're doing. I think that's incredibly

02:08:12.320 undervalued. We act as if it's a binary choice instead of why don't we see what else this area

02:08:17.600 has to offer? Maybe you just need to get on a different track. And again, I think that's the

02:08:21.280 brilliance of good coaches is they help personalize that environment so that the person continues to

02:08:25.460 be interested and continues to progress. And so the first thing I would try is varying up what

02:08:29.860 they're actually doing. And if they really want to quit, I'd say maybe keep a foot in if you can for a

02:08:33.920 little while. But then ultimately, if they really want to quit, then it defeats the purpose. If some

02:08:38.460 aspiration of life is to find happiness, then yeah. Yeah. And make it. And if they want to come

02:08:43.780 back, maybe they just needed a break, keep it available if possible. But if they quit and they

02:08:47.320 don't regret it, then what are you going to do? The proof's in the pudding. There are a couple other

02:08:50.940 stories in range that I wanted to ask you about. One of them is the story about the space shuttle

02:08:56.680 Challenger. I'll let you tell the story and then I'll bring my question to it.

02:09:00.860 I don't even know how to tell this story quickly. I'll set the stage. Everybody knows what happened.

02:09:05.500 Let's assume everybody knows the following. January 1986, space shuttle Challenger is scheduled

02:09:11.420 to launch, happens to be the first time there's a civilian on board. It also happens to be the first

02:09:16.100 time it's ever launched on a day that's that cold. And everybody knows what happened, which is

02:09:21.120 whatever, 73 seconds after liftoff, the O-ring failed to contain one of the liquid fuels. I can't

02:09:29.320 remember if it was liquid oxygen, but basically there was a spark, an explosion, and the rest is

02:09:34.760 history. You write about, one, the decision to launch that day. Two, the challenges of figuring

02:09:43.900 out what was the cause of that. Talk about both of those, especially the former, actually. And then

02:09:49.400 I'll tell you where I'm curious.

02:09:51.540 What I focused on was this emergency conference call the night before the launch when the weather

02:09:57.660 report came in that it was going to be an unusually cold day in Florida. And so engineers at Morton

02:10:04.080 Thiokol, which was the rocket booster contractor, and NASA, they get on this big conference call in

02:10:09.160 three different locations, a group of engineers, and they say, what should we do? Is this a problem?

02:10:13.420 Should we worry about it? The shuttle was supposed to be cleared for cool temperatures, but nobody

02:10:17.660 really knew because they had never launched below 53 degrees. And they were concerned.

02:10:22.120 As an aside, by the way, that O-ring was from the Apollo program. That's a great example of 1.00

02:10:26.620 engineering sort of shortcut, which is they never actually redesigned the O-rings for the shuttle.

02:10:32.600 They literally just took the O-rings from the Apollo project, which had a different spec,

02:10:37.120 and brought them over.

02:10:38.220 Maybe I should describe what O-rings are a little bit. It's like, it's a strip of rubber that runs along.

02:10:42.120 If you can picture the rocket booster, you know, it looks like a missile attached to the shuttle.

02:10:45.360 So the O-rings, it's put together in different vertical segments, and the O-rings run along

02:10:50.000 the perimeter of the missile-like rocket booster and seal the joints between pieces. And they have

02:10:55.620 to stay sealed so that they block the rocket fuel coming down the booster from shooting outside,

02:11:00.220 essentially. And there was concern that the rubber of the O-ring launched. There were forces that

02:11:05.620 moved the O-ring, that moved the metal pieces that were being sealed apart. And the O-ring rubber had

02:11:10.400 to expand immediately to maintain contact so that fuel didn't come shooting past. And the concern

02:11:15.940 was that the rubber would harden a little when it was cold. And so it wouldn't expand quite as quickly.

02:11:20.660 And so some rocket fuel would basically shoot through the wall of the booster. And that is

02:11:25.980 exactly what happened. And the question was that they were trying to answer is, should we be worried?

02:11:31.480 Because twice before, they had had cases where they saw soot on the wrong side of the O-ring,

02:11:38.860 which meant rocket fuel had gotten past it, but it wasn't catastrophic. One of those was when they

02:11:43.980 launched at 53 degrees, which was the coolest temperature they'd launched at. One was at like

02:11:48.480 75 degrees, which was one of the warmest temperatures they had launched at. And so-

02:11:53.000 And on this day, it was like 30-something, right? It was like 38 or something like that?

02:11:56.580 They were looking at it that it was going to be like 40, but it ended up being colder than they

02:11:59.600 expected even. And I think in retrospect, some of the temperatures of actual components were even

02:12:04.020 colder. There was like ice on there. So it ended up being colder than anybody expected anyway.

02:12:07.960 But their question was, in these temperatures, will the O-ring work or not, basically? And the fact

02:12:12.520 was that they didn't really... There were only two cases of this so-called blow-by when the rocket

02:12:16.600 fuel goes past. Again, one at the coolest temperature, one at one of the warmest temperatures. And the whole

02:12:22.020 shuttle was thought by the project manager to be cleared for lower temperatures anyway.

02:12:26.580 And so all of a sudden, they're having this last-minute meeting where they're saying,

02:12:29.100 like, how do O-rings work and when do they work? And the only real... They didn't have enough data

02:12:34.000 to answer the question. And one of the engineers was saying, we shouldn't do this because I inspected

02:12:39.700 the joints. And in the 53-degree day, there was lots of black soot behind the O-ring. So that means a

02:12:45.900 lot of gas got by and we're lucky it came back. And the 75-degree day, something else wants to happen

02:12:50.660 because only a little bit got past. I don't know what happened, but only a little got past.

02:12:54.220 And so we go colder. So he said, they asked him to quantify this. So what's the relationship between

02:12:59.520 temperature and gas blow-by? And he said, I don't know. I can't quantify it. All I know is colder is

02:13:05.520 away from goodness. That's how he put it. And they kept saying, quantify it, quantify it, quantify it.

02:13:10.180 He said, I can't. I've got photographs. That's it. And I think they're telling a story. And the fact that he

02:13:14.900 couldn't quantify it meant that essentially it was deemed an admissible evidence, basically, because they

02:13:19.680 couldn't quantify what the problem was. And so he made the decision to launch. The rest is history.

02:13:25.380 So the story is tragic on so many levels because it would almost be easier to accept a disaster like

02:13:31.560 that if nobody had seen what could have been done. That's what makes it so painful. But on a second

02:13:37.640 reading, what makes it even more painful is the way the questions are being asked, which is, were they

02:13:44.180 asking for him to assign a probability? Like, I don't even understand what the question was.

02:13:48.640 Quantify it. What does that mean?

02:13:50.300 Yeah. I mean, I think they wanted him to say at what temperature would it fail because the-

02:13:53.900 Well, it clearly had already failed. So really, this is a probabilistic question.

02:13:57.960 Exactly.

02:13:58.420 Yes. But again, that's not a linear question, right? That's an asymmetry question. That's the

02:14:03.540 question of so-and-so has an actuarial risk of having a heart attack of 5.4% in the next decade

02:14:11.600 should you take preventative measures? I don't know. I mean, it depends on your view of the

02:14:17.220 world. What's the upside? What's the downside? What's the risk of doing something about it

02:14:21.280 versus the risk of not doing something about it? Those are very asymmetric risks.

02:14:25.340 Yeah. And I mean, actually, if you read the Rogers Commission, which investigated the

02:14:28.620 Challenger disaster, Richard Feynman, you know-

02:14:31.040 My middle son is named after him, by the way.

02:14:32.760 Oh, okay. So Senna-

02:14:33.940 Senna and Feynman are two heroes of mine.

02:14:35.780 And your daughter named after-

02:14:37.200 Actually, she's named after her grandmother.

02:14:39.220 Oh, okay. Who was also, I'm sure, an exceptional person.

02:14:43.300 There was a point where Feynman was asking questions like this. Like, what were you trying

02:14:47.360 to get him to say? They say, well, he couldn't quantify his case. He didn't have data. And

02:14:50.580 Feynman says, when you don't have data, you have to use reason. And he was giving you reasons.

02:14:54.660 And there is a point at which they're trying to say, well, at this temperature, based on

02:14:58.480 this data, the chances of failure should have been like one in a, I don't know, a

02:15:01.760 fulfilling, some enormous number. And Feynman's like, that's nonsense. You had two cases

02:15:07.000 that had like a small failure here. And you're telling me it's basically impossible.

02:15:10.920 On a small, finite number of launches.

02:15:13.100 Yeah. Yeah. And so obviously there was a lot of cover your ass stuff going. And they even

02:15:17.160 knew that at the time. So there was this unusual on that conference call. Once Morton

02:15:21.700 Diacol said, well, our engineer is saying, don't launch. We're going to support that.

02:15:25.220 And then there was all this discussion and they went for an offline caucus where they

02:15:28.180 said, where they kept being asked for data, quantify your case, quantify your case, quantify

02:15:32.100 your case. And they said, we can't. Okay. I guess we agree to launch. And then

02:15:35.120 they were required to do a sign off that they didn't have to do in the past. So obviously

02:15:38.540 people were sort of in CYA mode at some level, or they felt protected by the process. But

02:15:43.540 yeah, I don't think they weren't the best questions that they were asking. I mean, I think they

02:15:47.580 wanted to know at what, again, at what temperature will we suffer catastrophic failure? And the

02:15:51.260 answer was that they didn't know, but that things didn't look good. And there was reason

02:15:55.340 to be worried.

02:15:56.800 Think about that question, even through the lens of the financial crisis, at what loan to

02:16:02.280 value ratio at, do you see a default? Or if you go through all the metrics, it's like,

02:16:07.120 it's a probability distribution. The only way you can really answer these questions

02:16:10.520 is you can't answer them this way, but the only way you can even get estimates at that

02:16:15.120 is to run simulations across distributions, assuming you even know how to predict the probability

02:16:21.880 of failure, which in that case, I would argue they didn't even have that.

02:16:24.900 And spatial, most complicated machine ever made, maybe not as complicated as the credit

02:16:29.440 system, but most complicated machine ever made, not whatever, 24 launches, they'd all

02:16:33.840 come back safely. And you're right, they didn't know. And in fact, when Morton Thiokol gave

02:16:38.840 its first recommendation, their initial recommendation was don't launch below 53 degrees because NASA

02:16:44.740 wanted a temperature. And I think the project manager thought that the whole shuttle was cleared

02:16:50.020 from 33 to 99. So they were putting definite boundaries on it. So they said, don't go below

02:16:54.980 53. And they said, well, what's your reasoning for that? They said, well, we've done 53 before

02:16:58.520 and it came back. And that actually totally backfired. So basically they set the lower

02:17:03.360 bound at what they'd already done and said, don't go below it. And so it backfired in the sense that

02:17:08.120 NASA had this very strong engineering culture, of course, that had worked great up till then.

02:17:13.100 And so he said, that's not science, that's tradition. That's not an answer. You've departed

02:17:17.460 from engineering. Give us an engineering story. Don't say just do what we've done before.

02:17:21.000 And I get that. Like as Mary Schaefer, a NASA engineer later said, perfect safety is for people

02:17:25.400 who don't have the balls to live in the real world. So you can't, it is a probability distribution.

02:17:29.220 You can't have perfect safety. But I think their attempt to be prudent backfired because it was

02:17:34.040 viewed as an emotional rather than a scientific case essentially. But as Feynman said, the data wasn't

02:17:41.140 there. So you have to start thinking about reason. And that doesn't mean maybe they would have

02:17:45.260 gone with the same solution, but they had to start thinking about this problem in a different

02:17:50.080 way than they were used to, because usually they did have the data they needed to make a decision.

02:17:53.680 And in this case, they didn't.

02:17:55.040 The other challenge of these case studies, because it's easy to just stop there, but we also don't

02:17:59.220 realize in the 24 launches prior and the God knows how many launches that followed it, including

02:18:05.960 one or more disaster, by the way, on re-entry, what's the denominator? How many times did somebody,

02:18:11.900 first of all, how many disasters were barely avoided? Maybe at least another one on the day

02:18:17.120 that it was 53. And could you ever have had a full consensus on any given day? This is a great case

02:18:23.340 study because it has so much, the richness of the data that follow are there, but we don't know if

02:18:30.420 every time one of these things goes up, there was also somebody in the room that said, and by the

02:18:35.100 way, with perfectly good reason, no way, no way, no way. And here's the reason. And if that's,

02:18:39.720 so is this the price we have to pay to live in an uncertain world? Or in retrospect, was Feynman

02:18:45.980 right? And this one should have been averted using reasonable engineering insight.

02:18:52.120 Yeah, I think, I think, and it's so easy to say in retrospect, of course, these people are under

02:18:56.260 incredible pressure. They've had all successes in the past. The astronauts know they're taking risks.

02:19:01.400 Everyone knows they're taking risks, but I do think there's a case to be made that this one

02:19:05.580 should have been averted, like voices were being raised. And in fact, I think one of the reforms

02:19:10.380 that came, you mentioned narrowly averting crisis before this, the 53 degree launch.

02:19:15.260 I think one of the, I spent a lot of time talking to the head of the rocket booster program while I

02:19:21.460 was reporting the book. And I think he was saying one of the changes that occurred was sharing

02:19:25.400 information like that with the astronauts themselves, because the feeling was, had they known

02:19:29.840 what happened with that O-ring, then they would have said, no, no, no, no, no, we're not ready to go.

02:19:34.720 You guys need to figure this out. And so I don't know whether that would have, or it wouldn't have.

02:19:39.320 That's an interesting idea.

02:19:40.220 But I think maybe, I think it's worth getting a different view when you're assessing that risk.

02:19:44.760 Was it Atlantis that, which was the one that burned upon re-entry, Columbia?

02:19:47.700 Columbia.

02:19:48.160 So totally different issue had to do with-

02:19:50.940 Same culture.

02:19:51.640 That's what I was going to ask you. Was there a cultural similarity in the screw up?

02:19:55.140 Totally.

02:19:55.580 Or was it a different, totally different type of miscarriage?

02:19:57.140 Cultural carbon copy. In fact, the investigation commission for the Columbia accident wrote in their

02:20:02.180 conclusions. This is so similar to the Challenger disaster, culturally, that we deem NASA not a

02:20:07.840 learning organization because they didn't learn from that experience. And what I write about is

02:20:11.940 they had this incredibly strong process culture, essentially, that was, they had this sign like

02:20:16.200 on the mission, on one of the rooms that said, in God we trust, all others bring data, which is

02:20:21.020 great. Like they were super rigorous. But when they would get into these situations where you didn't

02:20:26.340 have the data you wanted, then continuing to sort of ask for it and follow these very strict procedures

02:20:31.520 meant they really kind of constrained their thinking. Like with Columbia, there were engineers

02:20:35.920 who said, we'd like photographs of a part of the shuttle we think is damaged. And so they went and

02:20:41.760 asked the Department of Defense for those photos. And their superiors not only blocked them, but

02:20:48.180 apologized to the Department of Defense for going outside of the normal process for trying to

02:20:53.240 acquire things like that. And so in both cases, they kept sticking to this very rigid process.

02:20:57.980 We need an engineer case. We need a quantitative case. Or these concerns like they're not being

02:21:02.420 quantified and you're not going through the proper channels. And so it doesn't count.

02:21:05.600 Evidence and hunches kept being deemed inadmissible because they weren't part of the normal formal

02:21:09.800 process. And it happened in the exact same way. Both times, people suspected what was going to happen.

02:21:15.660 But because their concerns didn't fit into the normal procedural boxes, they were discounted.

02:21:20.420 No, we don't know the denominator. We don't know how many times people had concerns that

02:21:24.220 were completely unwarranted and would have thwarted. So therein lies the challenge.

02:21:28.680 And that's what Alan McDonald, who was the head of the Rocket Booster program, told me. He's like,

02:21:31.840 look, if we called off that launch for Challenger, people would be called, as he said, chicken littles.

02:21:37.080 Because you have to be willing to take risk in the space program and you don't really get credit

02:21:40.840 for not launching.

02:21:42.180 As I was about to say, let's assume they had not launched that day and they'd waited until the next day

02:21:47.000 and it was warmer and they launched. We wouldn't be having this discussion about it. You're never

02:21:50.940 a hero in that situation.

02:21:52.640 That's right. And I think there's one sort of thing that I didn't write about, but that I think has

02:21:55.600 been attributed a little wrongly, where some people have said there was pressure on Morton

02:21:59.440 Diacol, for sure.

02:22:00.780 More pressure because there was a civilian and this was more high profile? Or what was the...

02:22:04.220 No, for Diacol, there was more pressure because NASA had said they were going to open up the

02:22:07.760 Rocket Booster contract.

02:22:08.820 Got it.

02:22:09.160 But, so people have said, oh, that pressure, maybe that pressure was a problem, but Diacol

02:22:13.240 gave the initial recommendation not to launch. So they were okay not to launch. And then they

02:22:17.760 were pressed for the quantitative case. So I don't think that pressure was definitive because

02:22:22.280 they initially came with the recommendation of not to launch.

02:22:25.500 So speaking of stories in the book, this is different, but it's equally perplexing to me,

02:22:29.500 which is the firefighters and these guys who are doing these crazy jump rescue things. It's not that I

02:22:34.700 don't believe the stories. I do, but I can't. I'm trying to be empathetic to that situation and

02:22:40.900 say, where am I making a similar mistake in my life? So spend a moment explaining what that is.

02:22:45.540 Yeah. This comes from the work of a psychologist named Carl Weick, who writes a lot about what he

02:22:48.560 calls sense-making. Like how do people collectively make sense of a dynamic situation? And one of the

02:22:55.280 things he noticed when he was studying wilderness firefighting teams, hotshot firefighters who

02:23:01.080 hike in, try to dig trenches around wilderness fires or smoke jumpers who parachute into them

02:23:05.420 is that they do a great job. They're very reliable, but sometimes something unusual happens. Like a fire

02:23:10.620 jumps from one slope across a gulch to another slope and starts chasing them uphill. And when,

02:23:16.120 when unexpected things happen, sometimes they have trouble and sometimes they die. And when they die,

02:23:21.140 what he noticed was they tend, the ones who die tend to die with their tools, chainsaws,

02:23:26.020 drip torches, axes, whatever, hundreds of pounds of equipment. And the ones who survive or much

02:23:31.980 fewer have dropped their equipment and run. And in many cases, the hotshots or the firefighters will

02:23:38.160 refuse orders to drop their tools. And so I was going through like reports of some of these tragedies

02:23:44.720 and you'd see victim is 100 feet from safety, still carrying chainsaw and drip torch and backpack and

02:23:50.840 things like that. And even accounts of survivors would say they'd be running, they'd be looking for a

02:23:55.080 place to put their, they couldn't believe they were dropping their equipment because it was so

02:23:58.340 central to their identity as a firefighter. Norman McLean, who most famously wrote A River Runs

02:24:03.340 Through It, also wrote a book called The Young Men in Fire. And he wrote that being asked to drop your

02:24:07.720 tools is like being asked to forget that you are a firefighter because that's your whole group

02:24:12.560 identity and all your training is built on never getting rid of your tools. But an unfamiliar

02:24:16.480 situation, holding onto them kills you. And so Wyke used that as kind of a, an allegory for what he saw in

02:24:23.520 other usually highly reliable organizations like commercial airlines, where when things go

02:24:30.160 as expected, these very formal strict procedures work incredibly well. But sometimes having done

02:24:37.960 them so many times makes the organization rigid such that when things change and when it's obvious to an

02:24:42.900 outsider that they should drop their tools and run, they don't do it because they're so used to doing

02:24:46.740 one thing. And so what he was arguing for is how can we make training so that we have those reliable

02:24:51.220 procedures, but also so that people know they have to improvise and we can't train them exactly what

02:24:56.860 to do for improvising necessarily. Although now they do get trained to drop their tools, but who knows

02:25:01.060 what the other unfair situations are.

02:25:01.640 How many people had to die to figure that out? And again, the more powerful part of that allegory is

02:25:05.740 the, what's my tool, right? What are the tools that I am lugging around that are probably helping me

02:25:10.920 99% of the time, but 1% of the time, not only are they not helping me, but they could be

02:25:16.560 catastrophically wounding me. Yeah. I mean, have you thought about this in your own life?

02:25:20.780 Oh, for sure. In fact, when I told that, and this is not nearly as, it sounds like it has so little

02:25:25.720 gravity compared to this, but when I was talking before about taking a fiction writing class,

02:25:29.760 I was trying to think about the joy of it and think about new structure, learn new structure for writing. 0.59

02:25:34.820 But what really came out of it was this thing where it said to me, I'm using quotes in a stupid way

02:25:38.300 because I've been writing investigative magazine articles for the last couple of years where you and the

02:25:42.520 lawyers really want other people to explain stuff in their voice if they can, but that's not good for

02:25:47.300 this kind of book. And I didn't even think about that. It was not the right way to go about it until

02:25:52.100 I was like knocked out of it by doing something different. Again, that's not, that's not anything

02:25:55.880 on the level of tragedy. But some of the other things Wyke wrote about were doing commercial

02:25:59.660 airlines. They have these incredible, it's incredibly safe. It's pretty remarkable actually.

02:26:04.240 But when there are problems, it was usually the large majority of problems would be caused when

02:26:09.780 a situation would change and the flight crew would do the thing they were used to anyway,

02:26:15.100 even though to like an outsider would become obvious that they had to shift what they were

02:26:19.020 doing. So I think it's this sometimes paradox of expertise where if you haven't learned to do

02:26:24.180 some improvisation, then you kind of get stuck doing similar things over and over. You know,

02:26:27.360 you get like the typical, when all you have is a hammer, everything looks like a nail problem.

02:26:30.780 And I think there's plenty of that in medicine.

02:26:33.300 You could argue medicine is probably has more of that than anything else, given the degree of

02:26:38.880 specialization. But also, I think the tacit kind of, what's the word I'm looking for,

02:26:45.000 sort of the air of invincibility, not invincibility either, is the sort of the, I think part of it

02:26:49.860 stems from the privilege of medicine, which is like you can sit down with, you could meet a person for

02:26:53.700 the very first time in the emergency room, and they'll take their clothes off, and you're going

02:26:57.740 to examine them, and they'll tell you a detail about their life they've never told another human being.

02:27:01.960 There's just something about that that says, well, there's still something about this

02:27:05.260 profession we place so much faith in. And then by extension, those in the profession

02:27:10.740 start to project that faith on their own decisions. So I think that, coupled with the stakes,

02:27:17.720 coupled with the specialization, I think probably creates a fertile environment for exactly that

02:27:21.760 type of thinking. Now, I mean, I guess the last point I want to ask about on this particular issue is,

02:27:26.580 had those firefighters, going back to this particular example, been trained as lifelong firefighters

02:27:34.380 versus they had been sampled across a much broader group of people coming in? Is the point of the

02:27:42.760 story that if you'd come in to that role having been an accountant, working at Subway, just pick a

02:27:49.840 totally random distribution of people, would it be more intuitive to that group to have dropped their

02:27:54.940 tools? That's a suggestion. Yeah, that it would be. And in fact, what Wyke and McLean argued is that

02:28:01.240 in the first of the famous disasters called the Mangulch fire, a few people survived. And the leader of

02:28:08.320 the group, he ordered people to drop their tools. They refused the orders, essentially, except for the

02:28:14.000 two that did survive. And one of the things they pointed out was that he had a much broader training

02:28:19.960 base. He was used to doing lots of other things with tools, not only firefighting. And so they

02:28:25.500 thought that he was, we don't know for sure, but the argument that they made was that he had had a

02:28:30.700 much broader experience with tools. So he didn't think of them as like only these things that he

02:28:34.220 used in a certain way for firefighting. And so in fact, what he did to save himself was, so he's ordering

02:28:40.200 his guys to drop his tools. Most of them aren't. He realized he wasn't going to, he was far down the

02:28:44.840 gulch. He realized he wasn't going to be able to run away from the fire. So he actually lit a fire in

02:28:49.300 front of himself, burned the grass and dove into the ash and the fire burned around him. So he

02:28:55.380 improvised. And now people are trained to do that. The idea that you would stop a fire by lighting a

02:29:00.880 fire in the grass in front of yourself. And that worked. But so they made the argument that he was

02:29:05.220 much better set up to improvise because he was used to improvising with tools.

02:29:10.160 David, I could continue having this discussion for another couple of hours, but I want to be

02:29:14.440 sensitive to your time. And so before we wrap, is there anything that I feel like we've

02:29:19.280 only scratched the surface, frankly, of these two books? I actually thought we'd get through more,

02:29:23.680 but I guess that speaks to how long-winded I am when I ask questions.

02:29:27.060 No, you told me before we started that I could go on and be digressive. And I told you that I'm

02:29:31.100 naturally digressive. So I took you up on it.

02:29:32.860 It was a perfect, perfect, it was match quality at its finest. Anything we didn't talk about in the

02:29:37.480 last few minutes that you want a chance to sort of discuss?

02:29:41.640 Someone who became, that I don't get asked about much in the book, who became sort of a role model for me,

02:29:46.200 works not so far from here, a woman named Frances Hesselbein, who took her first real job at the

02:29:50.360 age of 54. And I'll keep her story short here. But basically, she essentially became, she became

02:29:56.660 a CEO of the Girl Scouts. And when she had one semester of junior college under her belt in her

02:30:02.840 entire life, but now has 23 honorary degrees, as she likes to know. And when she actually interviewed

02:30:08.520 for the CEO position, the people before her had had incredible leadership credentials. Captain Dorothy

02:30:14.460 Stratton was one, she started the Women's Coast Guard Reserve and was university dean. Another 0.99

02:30:18.040 was Cecily Kanin-Selby, prominent scientist and leader in industry and education. Frances Hesselbein,

02:30:23.660 one semester junior college, leader of one of 355 local councils of Girl Scouts. And again,

02:30:30.520 first professional job at age 54. And so she says, no, no, I'm not taking that CEO job. I'm never moving

02:30:35.720 out of Pennsylvania. She grew up in Johnstown. And her husband says, no, I'll drive to New York,

02:30:40.220 you can turn it down in person. And so they asked her, if you took over Girl Scouts, this is late

02:30:43.660 60s. Girl Scouts is in total crisis, free fall of membership and volunteers. And what would you do

02:30:50.080 if you took it over? And she doesn't want the job. She feels fine to say whatever she, to speak her

02:30:54.980 mind freely. And she says, throw out our sacrosanct handbook. I'd replace it with ones that appeal to

02:31:00.360 girls of different ages. I'd start working on diversity. If an indigenous girl near an ice 1.00

02:31:04.080 flow in Alaska opens a book, I want her to see herself in a Girl Scouts uniform. She just goes through

02:31:08.620 all this. I'd sell some of the underused campsites, even though it'd be painful. Get rid of some of this

02:31:13.360 homemaking stuff. Focus on, educate girls about sex and drugs and math and science and all this. 0.99

02:31:18.160 And she's like, well, that was fun, but I'm never going to hear from them again. Of course, she

02:31:21.220 comes back. I mean, this is actually the early 70s when she was interviewing. But so she gets the CEO

02:31:25.820 job and totally transforms the organization. Turns the cookie business into like a third of a billion

02:31:30.360 dollar business. Triples diversity representation, adds 130,000 volunteers, people she's paying in a sense

02:31:36.740 of mission, not in money, and basically saves the Girl Scouts. And she works every week. I'm sure 0.88

02:31:43.420 she's at her office right now because she works every weekday in Manhattan. She's only 103 and a

02:31:47.380 half. So she has a lot of, I'm sure she has a lot of place to go. But I think she meant a lot of

02:31:52.580 things to me. One, people can make an impact when they're older than they think they can. She's now

02:31:56.120 running the Francis Hesselbein Leadership Institute and teaching at West Point. Also, she never expected to

02:32:00.860 do any of the things she did. Every time she got offered an opportunity, she'd say like, no,

02:32:04.060 I'm not doing it. And then they'd tell her, well, then this Girl Scout troop is going to have to be

02:32:08.180 folded. Sorry. And she'd be like, fine, I'll do it for a month and would get to try it and be like,

02:32:12.240 wait, I actually love this. And that's everything she did. So she sort of short-term planned her way

02:32:16.780 through life. And she had two sayings that really stuck with me. One was leadership is a matter of

02:32:22.540 how to be, not a matter of what to do. I think that's a powerful thing to think about. So much about

02:32:27.600 leadership is being a good example, not having to know everything. Sometimes I think we have this

02:32:32.040 George Washington standing up in the boat crossing the Delaware, which I don't think that's an

02:32:35.820 accurate depiction. I think there's a real painting of him somewhere. And obviously he's

02:32:38.520 sitting down in the boat, but it's like this idea that they know everything ahead. They're

02:32:41.940 clairvoyant. It's like, no, they need to be a good example. And part of that is admitting that you

02:32:45.520 don't know stuff. And the other was her saying was, you have to carry a big basket to bring something

02:32:51.180 home. And she told me when she was at one of her first training events, some woman complained she 0.98

02:32:56.460 wasn't learning anything. She already knew all this stuff. And then this other woman told that saying to 0.95

02:33:01.480 Francis, and then it sort of became one of her mantras. And I love that saying, because I realized,

02:33:06.760 again, it was part of this, I don't want to keep coming back to this online beginners fiction

02:33:09.860 writing class I took, but it was, it was sort of the emphasis of hearing her say that they said,

02:33:13.460 sure, I can take a beginner's class. And I realized there's like no amount of beginner's classes I could

02:33:17.920 take that I wouldn't learn something from. Because if you go in open-minded, you'll learn something from

02:33:23.280 it. There was one day in my neighborhood, I live in DC now, where I noticed a bunch of

02:33:26.760 increase in the population of wizards in my neighborhood. And so I walked over to a nearby hotel,

02:33:31.120 noticed they were having like a Japanese conference, animation conference. Yeah.

02:33:34.460 And they had a beginning Japanese comics writing class. So I'm like, I'll sit in on that.

02:33:38.420 Probably not going to write a Japanese comic, but it's structure and it's story and narrative and

02:33:42.000 dialogue and all this stuff. You can't not learn from it. So it's just, just made me realize that if

02:33:47.700 you go into it with that mindset, you're just constantly learning something. And so that's an

02:33:51.740 approach I try to adopt. Two of those phrases really stuck with me.

02:33:54.860 And I am glad you told that story. That is a beautiful story. And I think it dovetails

02:33:59.600 perfectly into this idea of being a lifelong student, which I think, well, frankly, that's

02:34:05.520 sort of one of the things that's fun about a podcast. So basically there's an excuse to learn

02:34:09.160 a whole bunch of stuff. That's what I figured. I mean, you don't have to be doing this for any

02:34:11.580 professional reason. Yeah. Well, David, thank you so much. I really appreciate it. Thanks for coming

02:34:16.140 up to New York today. I appreciate it. Pleasure. I enjoyed it. Thank you for listening to this week's

02:34:20.700 episode of The Drive. If you're interested in diving deeper into any topics we discuss, we've created a

02:34:25.720 membership program that allows us to bring you more in-depth exclusive content without relying

02:34:30.480 on paid ads. It's our goal to ensure members get back much more than the price of the subscription.

02:34:36.100 Now to that end, membership benefits include a bunch of things. One, totally kick-ass comprehensive

02:34:41.380 podcast show notes that detail every topic, paper, person, thing we discuss on each episode.

02:34:46.680 The word on the street is nobody's show notes rival these. Monthly AMA episodes or ask me anything

02:34:53.040 episodes, hearing these episodes completely. Access to our private podcast feed that allows you to

02:34:58.500 hear everything without having to listen to spiels like this. The Qualies, which are a super short

02:35:04.280 podcast, typically less than five minutes that we release every Tuesday through Friday, highlighting

02:35:09.140 the best questions, topics, and tactics discussed on previous episodes of The Drive. This is a great way

02:35:14.240 to catch up on previous episodes without having to go back and necessarily listen to everyone.

02:35:19.060 Steep discounts on products that I believe in, but for which I'm not getting paid to endorse

02:35:24.520 and a whole bunch of other benefits that we continue to trickle in as time goes on. If you want to learn

02:35:29.700 more and access these member-only benefits, you can head over to peterattiamd.com forward slash

02:35:34.740 subscribe. You can find me on Twitter, Instagram, and Facebook, all with the ID Peter Attia MD. You can

02:35:42.600 also leave us a review on Apple Podcasts or whatever podcast player you listen on. This podcast is for

02:35:48.660 general informational purposes only and does not constitute the practice of medicine, nursing,

02:35:53.120 or other professional healthcare services, including the giving of medical advice. No doctor-patient

02:35:59.340 relationship is formed. The use of this information and the materials linked to this podcast is at the

02:36:05.160 user's own risk. The content on this podcast is not intended to be a substitute for professional

02:36:10.640 medical advice diagnosis or treatment. Users should not disregard or delay in obtaining medical advice

02:36:18.000 from any medical condition they have, and they should seek the assistance of their healthcare

02:36:22.600 professionals for any such conditions. Finally, I take conflicts of interest very seriously. For all of

02:36:29.300 my disclosures and the companies I invest in or advise, please visit peterattiamd.com forward slash

02:36:36.600 about where I keep an up-to-date and active list of such companies.

02:37:06.600 Thank you.

The Peter Attia Drive - March 09, 2020

#96 - David Epstein： How a range of experience leads to better performance in a highly specialized world

Episode Stats

Summary

Transcript