Hugo Scott-Gall: I am delighted to have with me today, David Epstein. David is a polymath and the embodiment of Range his new book is also the embodiment of The Sports Gene, his previous best-selling book, as he’s an extremely fast runner. He has worked with lots of high-performance organizations, UK premier league soccer teams, all the way to NASA.
At the heart of his work is, I would say, a fascination with high-performance; how does it come about and why? Welcome, David.
David Epstein: Thank you very much. I was a very fast runner, just to correct that introductions. Was. Now, I’m an avid jogger.
Hugo Scott-Gall: Well look, let’s get cracking. Here’s where I want to start. I think the central point you’re making in Range is that the current huge focus on specialization in careers, sports, music, science is wrong. That everything from Tiger Woods to tiger mothers says we should specialize, but you say the opposite, we should be generalists. Can you tell us why?
David Epstein: Well, I think the empirical reason is that when I examined research that actually looks at the development of elite performers, whether those are athletes, musicians, artists, technological inventors, it shows not the Tiger pattern; it shows, what I call in refence to Roger Federer, the Roger pattern. So, Roger Federer, we all know the story of Tiger Woods, is early specialization.
Roger Federer played a whole bunch of different sports. Mother was a tennis coach, refused to coach him be he wouldn’t return balls, normally. He continued playing badminton, basketball, soccer, long after his peers were focused on tennis, and that turns out to be typical for athletes who go on to become elite. When you look at the science that tracks their development, they have what sports scientists call a, “sampling period” where they try a bunch of things, gain a breadth of general skills, learn about their interest, learn about their abilities, and systematically delay specializing. And when they do specialize, they come to it with a much broader skillset, and that’s just the analogy for everything I’ve found in all these other domains I looked at.
And it’s not to say that specialists are unimportant, but I think that Tiger path is actually, really, the exception. We’ve been treating it as the norm. And I don’t think we’re in danger of undervaluing specialization, whereas we are in danger of missing out on a lot of the power of generalists.
Hugo Scott-Gall: I guess one pushback you could get to your prescription that actually breadth and delayed specialization is better, is that it’s quite risky. There’s quite a lot of societal or parental pressure on people to specialize. That seems a less risky path.
David Epstein: Absolutely. And again, the way I got into this was just looking at the evidence and being surprised myself, right? A lot of these stories we know, like the tiger mother on the first page of her book advertises the secrets to, I think she says, why Chinese parents raise so stereotypically successful kids. And one of the first things is, her children have to play either piano or violin. They can’t play any other instruments, and it turns out one of her daughters is getting really good at violin, and she’s supervising six hours of practice a day.
And the part that people don’t really remember is that her daughter ends up saying to her, “You picked this instrument, not me.” And quits. Which, actually, if you look at the research on this, that happens to be usually what happens when musicians are forced to specialize early, and they don’t have that sampling period where they can try a range of instruments, they usually end up quitting.
And so, I think some of the stories that we’ve, kind of, glommed onto are not really what we think when we look at them in a deeper way. That danger that you mentioned, you know, it feels dangerous. It’s psychologically upsetting to move around and to delay specializing. And that’s interesting you mention that because there’s research in the book, that I mentioned, where a pair of Harvard researchers are trying to figure out the habits of people who maximize what’s called their, match quality.
Match quality is a term economists use to describe a degree of fit between an individual, you know, interest and abilities, and the work that you do. And maximizing your match quality, or improving it turns out to be incredibly important for your motivation, for your rate of growth, and your overall success.
And what they found was that they people who tend to maximize their match quality are these people who instead of having some early plan where they specialize with some distant long-term goal, they, sort of, say, here’s where I am right now; here are my skills and interests; here are the opportunities in front of me; this is the one I’m going to try right now, and maybe a year from now I’ll change because I will have learned something about myself.
And they just do that over and over until they zig zag their way to a place where they can uniquely succeed and feel fulfilled. And the project, the researchers named it The Dark Horse Project, which wasn’t the name when they started. But it was because all of these incredibly successful people from this – and fulfilled, that was one of the other requirements from these diverse domains, you know, from master sommeliers, to investment bankers would come in and say, well, I did all this other different stuff, and I came in through the side door, so I’m an oddball. So, don’t tell people to do what I do.
And that’s what, like, 90% of the people were saying. They all viewed themselves as dark horses. So, I think this thing that we think is a real exception is actually not, but we still have this neurosis about it.
Hugo Scott-Gall: Yeah, and maybe the most extreme example in the book is the story of Vincent Van Gogh. You made the case for him being a real browser trying lots of things, and it’s interesting to me that, obviously, this is retrofitting that Van Gogh would have scored below average on Angela Duckworth’s Grit Scale. So, can you resolve the contradiction between grit, which let’s just say is dogged perseverance, and browsing before finding your calling because certainly, grit has held up as one of the key quality and attributes that we should all display to show that we’re high performance individuals?
David Epstein: Right, and the reason I used Van Gogh is because he’s one of the most well-known and beloved people, not just artists, I think, or cultural figures, and created some of the most valuable objects in the history of mankind, and was at the age of 27 when he picked up the book, The Guide to the ABCs of Drawing. He’d had five careers before that. And he had an incredible work ethic; he was an incredibly hard worker. He would dive into whatever he was doing, assuming that working really hard would lead him to success, and then he would fail and totally change directions. He did that over and over and over until he found something that he was really good at.
So, the reason, and I had one of his Pulitzer Prize winning biographers fill out the so-called Grit Scale for him. So, we all know this concept of grit, but we don’t really know what it’s based on. All right, it’s based on a 12-question survey where six of the questions test your perseverance, they ask, “Do you bounce back from setbacks?” Those sorts of things. And the other six questions test your, “consistency of interests.”
So, when Van Gogh’s biographer filled this out, he scored him very high in work ethic and low on consistency of interest because he was always bouncing around, even once he became an artist, he bounced around within that. So, when grit awards these points for consistency of interests, the studies that that comes from are very short term. And I think that’s a problem for how we extrapolate it to the rest of the world.
So, the most famous grit study of all was done on cadets going into the US Military Academy at West Point, and it was on them getting through what’s called, beast barracks. It’s this six-week orientation where they come from high school, and now they have six weeks where they really get indoctrinated into the military, and face physical and emotional challenges, and things like that. It’s a fragile time for them. Some of them realize they don’t want to be in the military and quit. Most of them make it through; the vast majority.
But what Angela Duckworth and her colleges found was that the Grit Survey was a better predictor of who would make it through than were other more traditional measures, like their test scores, their athleticism, and other things like that. And so, that’s good to know. But most of them get through, first of all, and secondly, that study looks at a very small group of highly preselected people in a six-week program.
So, the goal is very narrow, it’s very well defined for them, and it ends in six weeks. Okay? And that is not really something that we should extrapolate to the rest of life because if you take a bigger picture look and follow West Point cadets who, those gritty cadets who get through beast barracks, and get through the military academy, about half of them drop out of the army the day they’re allowed to.
And the reason that these cadets are leaving is the more that they’ve learned about themselves in their early 20s, the more skills they’ve learned, and the higher potential they are, the more they can transfer those skills to other areas of work, and the more opportunity there is for lateral mobility in society, in a way that there didn’t used to be where it was smarter to specialize earlier where you were facing the same repetitive challenges over and over in your work, but that’s not the case now.
And so, I think it’s really problematic to extrapolate grit way beyond any environment in which it has ever been studied because the environments are these situations like beast barracks, like the finals of the national spelling bee where you’re already starting studying people who are already in the finals.
And so, I think we’ve extrapolated it in a way that contradicts other research about how people find and succeed in the best careers for them, which has a lot to do with maximizing their match quality. And the way you maximize your match quality is by trying different things, reflecting on it, and them moving forward accordingly.
Hugo Scott-Gall: Sure. I just want to track back a bit and talk about the important distinction you make between kind versus wicked environments. Early specialization in things like golf and chess work because they are kind environments, but you make this important distinction between kind and wicked. And so, could you expand a bit on that. I think that would be really helpful.
David Epstein: Yeah, so these are terms coined by psychologist Robin Hogarth, and this was based on a disagreement between two famous researchers of expertise and judgement. And one was Danny Conoman, of course, the Nobel laureate who studies cognitive bias, and the other, Gary Klein. And what Conoman found was that lots of experts didn’t get any better with experience. And sometimes, they got no better but more confident. And Klein found that experts did get better. And what they realized was that, actually, when they came together, is it depends on what kind of environments those experts were learning in.
If they were in a kind learning environment, which is like golf. Okay, in a kind learning environment, all the information is visible, easily available, the next steps are clear, the goals are clear, your feedback every time you do something is immediate and perfectly accurate, so you get better just by doing things more, and more, and more, like golf. People wait to take turns; all those sorts of things.
Wicked environments, information is hidden, people don’t wait for each other to take turns, the next steps may not be clear, you may or may not get feedback all the time, it may be delayed, it may be inaccurate, it may even reinforce the wrong lessons, and in those areas, people that are too narrowly specialized will not get better, but they will get more confident. In some cases, they’ll even get worse.
This famous example of a famous physician in New York City who was renowned and became rich and prominent because he could diagnose typhoid weeks before a patient showed any symptoms whatsoever, or he could predict that it would come, and he would do that by palpating their tongue, or feeling around with their hands, and as one of this colleagues later pointed out, he was a more productive carrier than Typhoid Mary just with his hands because he was passing it, and the feedback was reinforcing the exact wrong lesson because he kept making correct predictions.
Now, most of us aren’t in that wicked of an environment, but most of us are in an environment much more toward the wicked end of the spectrum than golf.
Hugo Scott-Gall: So, is the key thing to take away from that that it’s humans that have the ability to integrate broadly an unmapped territory? If you map the real world into quadrants of specialized versus broad and kind versus wicked, it’s more likely, and maybe increasingly likely that you’re going to experience in your life a lot more broad and wicked, than you are specialized and kind?
David Epstein: At least in the work world. So, if you count everything we do every day, there’s plenty of kind environment, right? This can be as simple as things as you use a sink in one building, you’re going to know how a sink in another building works, right? But those are things we take for granted. When we’re thinking about the much more complex world of work where we’re actively trying to get better, it usually is wicked. There’s some information hidden; feedback isn’t automatic; it’s rarely fully accurate.
And in those situations, having an increasingly narrow view as you get more specialized, you’re basically taking a more, and more narrow view of a large system. And that can lead to some really perverse effects because you’re no longer seeing the system. You’re only having a keyhole view into this much more wicked environment.
Hugo Scott-Gall: I want to take this a step further and say, are we moving here into humans versus machines that rules based activities are best done by machines, and therefore, maybe activities like investing can be done by machines. But as the saying goes, all the danger’s in the past, all the vanity’s in the future, and if the future looks like the past, i.e. it’s kind, it’s fine for machines, but if the future is going to be different and strange, i.e. wicked, then maybe humans are actually better equipped to deal with an uncertain or different future versus the past.
Would that make sense to you as a, kind of, maybe I’m looking for things that aren’t quite there, or reinforcing it, but it does seem to me that an uncertain future or a different future is more fertile territory for humans because they can deal with the wicked environment more easily than a machine could.
David Epstein: Absolutely. And I think you see that in – I, kind of, discuss this in the first chapter of Range, and you see this in the things that computers conquer early, right? So, chess, 1997 when Deep Blue beat Garry Kasparov, the cover in Newsweek for that match was, “The Brain’s Last Stand.” Right? So, as if the computer wins, then what else do humans have to contribute because chess has been viewed as this epitome of human intellect; the embodiment of intellect. If we want to give someone a compliment on their strategy, we say, they’re playing chess.
But it turns out, chess is, like, a perfect activity for computers. It’s based on pattern recognition, like recurring patterns, and there’s an enormous store of previous data, and it is incredibly rule bound, right? Every piece can only move in very certain discrete ways, and so it’s rule bound, it’s discrete, people wait for eachother to take turns, and winning is based on seeing recurring patterns. And so, in that area, computers have made exponential progress. Like, now the free chess app on your phone can beat Garry Kasparov.
If you go to a somewhat regular but slightly less consistent environment like driving, so if you think of self-driving cars, there’s been huge progress. Huge. But challenges still remain. And some of them are formidable. So, that’s an area where there’s tight regulation, most of the things that happen are, basically, recurring patterns, but there are some irregularities that have been a challenge in that area.
Then if you go to the far end of the spectrum, something like cancer research, right, where IBM Watson has been such a titanic flop that some of the AI researchers I talked to for this book were worried that it would damage the reputation of AI in healthcare generally because it’s been such a flop. And as one of the oncologists told me, the reason why Watson was great on Jeopardy and poor in cancer research is that we know the answers to Jeopardy. All right, so if we already know the answers, and it’s a recurring environment, and based on recurring patterns, then I’m not sure humans always have a ton to add in those situations.
But when it’s not like that, when it’s a more open-world problem, we’re trouncing machines. And I think ultimately, we can outsource some of the skills that are normally learned via specialization to machines, to liberate us to do the broader conceptual thinking and integration that we are uniquely good at/
Hugo Scott-Gall: That definitely makes sense. I’d like to talk a bit about foxes and hedgehogs, and Philip Tetlock. You spent time with him, you know his work on superforecasting inside out. I think it’s very relevant here. I see foxes and hedgehogs; generalists and specialists as being from the same school of thought. Can you talk a bit about what you took away most from his work, and then maybe describe who are the foxiest foxes and why?
David Epstein: Yeah, so this Phil’s work is, basically, the core of chapter 10 of Range, and this started in the ’80s where he was a very junior member at the national research council’s committee on American/Soviet relations and noticed that these renowned experts in the field were making predictions about what would happen in the world, and that they were often perfectly contradictory, but they were making them very authoritatively. And so, he decided to start a project to study expert forecasts.
Most forecasts that we see, like on TV and in the financial page, like, now I read the business pages and just laugh because it’s like somebody predicting that housing will go in one direction with no timeline. They don’t give a timeline, they don’t say how much up or down it’s going to go, so it’s not even a prediction. And I was reading a Paul Krugman column recently where he was making some geopolitical prediction, and he said, “My answer is definitely maybe.” And I’m like, what percent chance is definitely maybe?
But anyway, Tetlock had to get specific forecasts, so people had to give specific probabilities of events happening with specific deadlines, and he had to collect enough forecasts that he could differentiate just lucky and unlucky steaks from skill.
And so, his project ran 20 years, took about 82,000 different forecasts, and what he found was that the credentials that make people visible experts had nothing to do with how good they were at forecasting. And what had everything to do with how good they were at forecasting was how broad their information consumptions was, how broad their curiosity was, how broad their mental models were, how many different perspectives they tried to aggregate, to use his language, as before they forecast, and nothing to do with their specific areas of expertise.
In fact, the worst forecasters were the most highly specialized individuals who’d worked on one problem of through one framework their entire career, and those people got worse as they accumulated more credentials. So, first PhD and many things beyond that, they actually got worse. And the suggestion is that their view was so narrow, they had this singular view on things they had worked on, that they could always find enough information to fit that view because they knew everything in their little keyhole, and could bend everything to fit their singular view, even though that had nothing to do with the world.
And so, I think he unsettled a lot of people and the US intelligence community took an interest in his work, pit some forecasters against intelligence analysts who have access to classified data, and he basically, just picked a bunch of people from the general public who have really wide ranging reading habits and are really curious, and not very dogmatic, and they destroyed the analysts with access to classified information. And these people were just members of the general public who did not have access to classified information, so it was even more extraordinary in that way. And those were what he called, the foxes.
Hugo Scott-Gall: That’s an amazing and profound piece of work with all sorts of conclusions that are coming off of it. I remember seeing in a CIA paper from, at least, 20 years ago, there was a great chart, which was a number of items of information versus confidence and accuracy, and as people gained more and more items of information, their confidence went up in a pretty linear way, but their accuracy flat lines at around 20%. Which I think is very interesting. It’s very interesting for the financial world in that there’s definitely increased confidence, or increased security from having lots of information. It shows you’ve put the effort in, and who doesn’t like a hard worker.
But I think there’s an inherent fragility behind the strength, which is, maybe a hedgehog; a specialist lacks the confidence of a fox, or a generalist, who is able to make more decisions earlier in a process that can be more accurate. Which isn’t to say they don’t update their views, but they just need less to go on. One of Jeff Bezos’s letters from Amazon says, “Look, you’ll never get all the information you want. You have to make decisions earlier than you would like, but that’s just the reality. You probably make decisions with 60 to 70% of what you’d like, but that’s just how it is.”
And so, I wonder whether, maybe, incentives in some forecasting industries, which would include financial services; it would include intelligence, is that source of, hey look, I’ve done all this work; I’ve gathered all this information, so you can’t be too hard on me if I’m wrong. I’m going to hide behind that. Is that a sort of a hedgehog specialist’s reflex behavior?
David Epstein: That’s interesting, and I think you brought up two separate issues. So, I’ll try to address them a little bit separately. One is, so the hedgehogs are drawn from this philosophy, I’d say, but the hedgehogs, as Tetlock says, “Know one big thing.” So, those are the narrow experts. And the fox know many little things, are the generalists. And the foxes, actually, were not more confident. They actually tend to be less confident. And I wouldn’t say that they systematically get less information necessarily, it’s really about their habits of mind, and what they do with information.
And that gets to the second point. What the hedgehogs tend to do, is if they have to predict a scenario, they drill into all of the little specific details of that exact scenario. They tend to get worse with more information, and the foxes tend to get better even if they’re dealing with the same scenarios. So, while they’re drilling into that information, the foxes take a much broader view and start looking for other similar scenarios, like, in history. Instead of drilling into that scenario, they go broad and look at other things, right?
One of the prediction questions, like these were hard questions was, will Greece have a currency conversion when it was having its debt negotiations with the eurozone? And this hadn’t happened before, maybe it was as if Greece was going to leave the eurozone, whatever it was. No one had left before, so there was no precedent.
And what the hedgehogs did, is they drilled into all the specifics of who are the leaders engaged here, and all these little specific things about Greece and the scenario, where as the foxes were much more likely to say, okay, yes, this is unprecedented, nobody’s left before, but lots of countries have left international agreements, and alliance, and coalitions, and there have been a number of forced currency conversions, and so forth, so they basically went looking for analogies in other areas.
It turns out that that inside view tends to be extremely inaccurate. Whatever scenario you start investigating with the inside view, you will just deem it more and more likely as you get more information.
And so, the foxes always take that outside view instead of focusing too much on a particular detail, they look much more broadly for information, and that, kind of, sort of, in a way, makes you think like a statistician. You’re basically looking for base rates of what usually happens in scenarios like this. And most things aren’t 100% unique.
So, that’s just an example of how that foxes take this broader, so called, outside view, whereas the hedgehogs drill deeper, and deeper, and deeper into all the particular details of the scenario, and that turns out not to be a good strategy for forecasting. And even though I think it’s the intuitive strategy that most of us have for everything we want to evaluate.
Hugo Scott-Gall: Yeah. So, here’s a question I just asked myself through the whole of the book, which is, can you teach conceptual thinking? One of the studies you refer to in the book says that, actually, GPA scores serve as a poor predictor of performance in conceptual thinking. So, can you actually teach mental models: mental frameworks; conceptual thinking? Is this where learning fire; desirable difficulties comes in? I thought that was a very good phrase you picked up on.
David Epstein: Yeah, and I think you can. So, desirable difficulties is a term coined by Robert Bjork, a psychologist who researches learning, essentially. And what it means is difficulties in learning that you want because they actually make learning better. And it turns out that the way that we usually gage our progress in learning is by tests of immediate progress, essentially. And what you really want is to set up certain types of obstacles so that learning becomes harder. But that instead of teaching you, so called, using procedures knowledge, where you learn how to execute certain procedures, you learn the more conceptual knowledge of matching a strategy to a type of problem, basically.
So, I don’t know how in the weeds you want to get about that, but basically, there are certain types of desirable difficulties you can use that cause someone learning the same material to learn conceptual thinking, as opposed to learning just how to execute procedures for the scenario
Hugo Scott-Gall: I think you made the point well, which is if you’re learning fast, you’re less likely to retain. If you’re learning slow, and indeed if you’re being tested in the same timeframe, it can lead to really quite misleading results. I thought that was a very very important point.
David Epstein: And not only more likely to retain, but even more important than that, the slow learning techniques make you more able to live in a wicked environment where you have to apply the information to scenarios you’ve never seen.
So, to use one example that I use in the book, I’m training these simulations training naval commanders to respond to certain types of threats, one type of group will train on a certain type of problem over, and over, and over, and over this certain threat, and they learn how to respond, and then they move to the next problem over, and over, and over, learn how to respond, and then the next, and so on, and they get really good because they’re practicing the same thing over and over.
The other group gets different scenarios every time, and they’re really frustrated because they’re not getting better because they’re never seeing the same thing. But then, when you bring those people back and face them with scenarios that neither of them have ever seen before, the frustrated group destroys the group that was practicing the same things over and over. Because instead of just learning how to execute the procedures by doing the same thing over and over, they’re forced to try to match types of strategies to types of problems, and that’s how you build these more general conceptual frameworks that you can apply.
And so, when they both see something they’ve never seen before, only one group has that more conceptual, flexible knowledge. And that goes to this classic research finding that can summed up as, breadth of training predicts breadth of transfer. So, the more broad your training, the more able you are to transfer your knowledge to scenarios you’ve never seen before.
Hugo Scott-Gall: Having completed your second book, what would you have done differently with your first book, The Sports Gene, which talks a lot about – it was all about where sporting performance, sporting efforts come from, it isn’t necessarily what you think. You take a slightly different view from your friend Malcom Gladwell who has taken a lot of the research, around 10,000 hours, and said, look, 10,000 hours, he doesn’t quite say this, I’m fairly paraphrasing, but with 10,000 hours you can be anything you want.
And you’re saying, well, actually, that’s not true. Unless you’ve got particularly long Achilles tendons, you can’t be a high jumper. Unless you’ve got a long trunk, short legs, but very wide arm span, you can’t really be a high-level swimmer. You’re saying, look, nature does matter. It isn’t just nurture.
Having written this book about generalists versus specialists, what would you, if you were writing The Sports Gene again, what would you put in? What would you emphasize more? Do you think the equation for successful sports people is maybe a little different from what you had down then?
David Epstein: A little bit. The intro to Range is called Roger versus Tiger because it, sort of, grew out of The Sports Gene. And by the way, Gladwell and I were just at the MIT Sloan sports analytics conference on a panel, and it’s on YouTube, and on the end of the panel, he says, “I changed my mind about something. I wrote about the 10,000-hour rule,” obviously, a lot of deliberate practice is important, that’s totally uncontroversial, but he said, “I assumed that implied that early specialization was the way to go, and I’ve changed my mind.”
And so, you can go see that on YouTube, if you want to see him talk about that a little bit. But I think I would have gotten into that a little more. I talked a little bit here about match quality; the importance of matching your abilities and interests to what you do, and that turns out to be incredibly important for people’s careers, and I think I would have liked to get into that a little more.
And I also think I would have gotten into some of this breadth or training predicts breadth of transfer work because I find it fascinating, and I think it’s some of the reason behind why athletes that have a sampling period early on then seem to pick up sporting skills better and more quickly as they go forward.
Hugo Scott-Gall: You visited and asked to speak to elite sporting organizations all around the world. What are the most frequently asked questions? What is it they’re looking to you to answer for them?
David Epstein: Oh, that’s a good question. I mean, it depends what level they’re dealing with. So, when I wrote The Sports Gene, first it was they were wondering about what type of physiology should they maybe screening for, and did they have to start getting interested in genetic testing, and all those sorts of things like that. Since I’d been through all the research in genetics, should they be screening, and the answer, basically, is that they should not be using genetic screening.
But then, as I started writing more about athletic development, it started being more about, how can we arrange our developmental pipelines to try to maximize more of our individuals? And so, that is more relevant to Range, and the answer in many cases, so the UK, for example, when it was awarded the Olympics decided to make sports science on of its competitive advantages, and the main thing that it did to really turn around UK sports, which had been having some mediocre Olympics for quite a while, was diversifying the talent pipeline. Right?
The woman, Chelsea Warr, who helped design this called it – she said, “We have to make room for fast risers and slow bakers.” To use her terms. She said, the fast risers, everybody’s looking for them, the kids who are much more bigger, much more developed, or whatever, but we now know that most of the people who go on to become the best athletes are the, so called, slow bakers, who you don’t necessarily pick them out as so great at first, but they develop, maybe they zigzag a little in their development, or they come along a little slower; maybe they physically develop a little slower.
And so, recreating the pipeline to make sure that you don’t accidentally kick those people out is a huge revelation. So, it’s fixing the pipeline, tweaking the pipeline to make it more amenable to those slow bakers, who we know often go on to become some of the best in the world is what I get asked a lot about now.
Hugo Scott-Gall: And slow bakers, that’s the core of the book, right? That’s what you’re advocating.
David Epstein: Yeah, and it’s part of what I got interested in. And it’s not to say that they don’t specialize eventually, right? I do think Roger vs. Tiger model, but of course, we all specialize to some degree or another at some point or another. It’s a question of how we get there, and what kind of view and skills we bring to it once we get there.
Hugo Scott-Gall: If I could just round off with one final question, which is really about your process. You’ve written two fascinating books that are very ambitious in scope, and they’re really quite different, you must have to do, and thumbing through the footnotes, you do do a ton of research. Where do the ideas come from, and does the end product look like the original idea? How do you change your thinking, and as per superforecasting how do you revisit your thesis; refine your thesis? How do you just gather all this information? Is it the loneliness of the long-distance research, and not the loneliness of the long-distance runner?
David Epstein: It is lonely, and it doesn’t surprise me that there’s such a vibrant literature among long-distance runners like myself because I think you have to be able to spend a lot of time in your own head, frankly, to do either one of things: writing books or long-distance running. So, for me, both my books, no, they look nothing like the book proposals. And Range looks a little more like the book proposal. Maybe half like the book proposal that I started with, whereas The Sports Gene looks nothing like the book proposal I started with.
In both cases, for the first year, I just tried to read ten scientific journal articles a day every day for the first year. Don’t write. At least I don’t write finished words. That’s my expansive search where I cast my net really wide, and it’s very inefficient because sometimes I’d go down a rabbit hole that I’d come up a week later and say, how did I ever think I was going to write about that? But I think accepting that inefficiency is also what allows me to find things that other people aren’t finding in many cases.
And I keep this thing going called a master thought list where I put down thoughts or sources, or whatever on a document, and as they start coalescing around a topic, I move them toward one another, and when there’s a bunch, I put a tag over that information, like the name of whatever topic it is, and a bunch of words that I think I would search if I wanted to find it. And then as I generate more of those tags, I start moving them, the relevant ones toward one another. And it, basically, ends up like a movie storyboard. And that becomes my outline.
Of course, it’s, like, 60,000 words long, so it’s, kind of, unwieldy, but I go in and I take these ideas, and then try to falsify them, right? Whether they’re mine or someone else’s idea, I say, here’s my intuition. The Sports Gene started with why baseball hitters can’t hit softball pitchers, and I started with the hypothesis, the first one I could think of was that they don’t have fast enough reflexes. So, then you go and test that, and it turns out it doesn’t really have anything to do with their reflexes at all, but that’s what I do.
And so, Range came out of my first debate with Malcom Gladwell where, at the MIT Sloan sorts analytics conference where I knew he was going to argue for the importance for an early head start in specialization for the development of athletes. And so, I said, okay, if that’s the hypothesis, I’ll go look at the data on the development of athletes, and it showed that the future elites tend to have the sampling period in delayed specialization. I said, if this is your hypothesis, it does not comport with the data. And so, that’s, basically, what I did.
And I expected Range – the original proposal was called Roger Versus Tiger because it was going to contrast those developmental models of Roger Federer and Tiger Woods, and say, when is it better to be a Roger and when is it better to be a Tiger? But I found do much more Roger than Tiger that I decided to focus the book on that instead.
Hugo Scott-Gall: Well, David, if I could just say thank you very much for giving us so much time. That was fascinating. It always is. The first time we met was actually in a bar in New York, and you invited me to join your running club, which was you and Malcom Gladwell. I did my research on both of you and concluded that me joining a running club would be an utter disaster for me. But I regret it because I would have been fitter, and I would have learned an awful lot more.
David Epstein: And do you know, to his credit, we became running buddies. The first time we ever met was for our debate, and then he invited me to run the next day, and then we started talking about this stuff on our own time, so it was nice to have productive debate that way.
Hugo Scott-Gall: So, over 800 meters, head to head, who wins? You or Malcom?
David Epstein: Over 800? Me. If you were talking about us in our prime, kind of thing, and our best performances, me, but over a mile, him.
Hugo Scott-Gall: Okay.
David Epstein: He’s a borderline world-class miler for his age group, now. And he was a Canadian provincial champion, so I would want to get him in the shorter distances; he would go longer.
Hugo Scott-Gall: Yeah. Yeah. You’re both the sports gene, you’re both 10,000 hours, your grit, and your growth mindset; you’ve got everything.
David Epstein: I appreciate that, even though I didn’t score well on the grit scale because my interests change from time to time, as you can tell from my book topics
Hugo Scott-Gall: True. True. David, thank you so much again. It’s always a pleasure. And thank you.
David Epstein: Pleasure’s mine.
