In his second appearance, Nate Silver joins the show to cover the intersections of predictions, politics, and poker with Tyler. They tackle how coin flips solve status quo bias, gambling’s origins in divination, what kinds of betting Nate would ban, why he’s been limited on several of the New York sports betting sites, how game theory changed poker tournaments, whether poker players make for good employees, running and leaving FiveThirtyEight, why funky batting stances have disappeared, AI’s impact on sports analytics, the most underrated NBA statistic, Sam Bankman-Fried’s place in “the River,” the trait effective altruists need to develop, the stupidest risks Tyler and Nate would take, prediction markets, how many monumental political decisions have been done under the influence of drugs, and more.
Watch the full conversation
Subscribe on Apple Podcasts, Spotify, or your favorite podcast app to be notified when a new episode releases.
Recorded July 22nd, 2024
Read the full transcript
Thank you to an anonymous listener for sponsoring this transcript.
TYLER COWEN: Hello, everyone, and welcome back to Conversations with Tyler. Today I’m in New York City, chatting with the great Nate Silver. It is late July 2024, and Nate has recently risen in status, you might say. He has a new book out, which I thought was excellent, informative, fun on every page: On the Edge: The Art of Risking Everything. Nate, welcome.
NATE SILVER: Thank you, Tyler.
COWEN: Now, if we simulated the world a thousand times, in how many of those scenarios would you end up more or less where you are today?
SILVER: [laughs] It’s, of course, a question I asked Peter Thiel and other people in the book.
COWEN: Your father was a political scientist, presumably very smart. Got you connected to politics.
SILVER: This is a hard question. I feel it’s unfair that you’re asking me this question that I asked other people, but I think I would have wound up somewhere in this vicinity 20 percent of the time, or maybe less than that. I don’t know, because I had this big breakthrough in the 2008 election, which really did involve a lot of happenstance in certain ways.
I had been an online poker player, and basically, the US government shut down online poker. Technically, they shut down payment processing for online poker. That got me very interested in the 2006 midterms because I wanted the people who had passed that law to lose their seats in Congress.
Meanwhile, I lost my source of livelihood. I couldn’t play poker and press buttons [laughs] for 24 hours — not 24 hours — for 40 hours a week. So, I founded FiveThirtyEight just on a lark, and that changed everything. Yes, there’s a lot of circumstance and luck, and there are probably many ways that life could have unfolded where I was unhappy at some consulting job, or something tragic happened, or whatever else.
COWEN: Poker players — they’re ornery. They’re individualistic. You don’t always want to hire them, and you’re one of them. Isn’t it, if not inevitable, likely you would have ended up as some kind of independent? Everyone knows about sports.
SILVER: Yes, because now I’m doing three or four different things, which in some ways makes it feel like I’m living out multiple versions of the simulation anyway. [laughs]
COWEN: All at once.
SILVER: Yes, because I’m diversified in different ways. There might be worlds where I had won a big poker tournament early on, or something like that, but yes, I get pretty unhappy when I’m bored. I also can focus and work pretty hard, so I think I would have wound up in interesting places a lot, but I feel very blessed and lucky to have wound up in this particular place.
COWEN: What do you maximize?
SILVER: I don’t know. In the short run, some combination of . . . Look, this year, it’s easier to answer that because you have a short or more medium-term . . .
COWEN: Right, everything’s thrown at you. You’ve got to get through the day, simply.
SILVER: Yes. Look, I’m 46. There’s some extent to which I feel like this year is important as far as earning a fair amount of income — because it is very cyclical, with election cycles and other things like that — of having good options for the second half of my life, basically.
I found a lot of energy recently, this year. I probably work fewer hours than I once did but probably work much more efficiently, have a little bit more balance. But yes, I am trying to get as much productive work done as I can over the next four or five months until the election happens. I’m trying to build the newsletter Silver Bulletin in the long term.
I’ve matured a little bit. I tend not to work most evenings, for example. We were talking about travel before, right?
COWEN: Yes.
SILVER: I do my share of travel. I have poker tournaments, which are fun and stressful but also a lot of hard work and are an escape in different ways. I want to get to a point where, for the rest of my life, I can do intellectually challenging and stimulating work and experience different things. But this year, there’s a little bit of hustle required to help ensure that in the long term.
COWEN: There’s a Steve Levitt paper you cite in your book that indicates people make better decisions when they flip a coin to avoid status quo bias. (A) Do you believe that paper? And (B) looking forward, will you be flipping any coins?
SILVER: No, I literally will flip coins for like, “Do we want to go get Italian food or sushi tonight?”
COWEN: But a big decision. Bigger decision, even just, “Where should I go on my next vacation?” You could randomize that. Do you do it?
SILVER: A vacation location might be up to the max of what I would randomize, I think.
COWEN: Not something really important.
SILVER: I tend to be pretty analytical about really important career and life-choice decisions, I think.
COWEN: So, you don’t think you have that much status quo bias?
SILVER: I’m sure I do, but look, I’ve done a lot of different things, and I’ve bounced around quite a bit. I’m inherently restless. Some of the common personality flaws that . . . “Flaws” is a jaded, loaded word. Some of the common personality traits a lot of people have, I think maybe poker players over-index a little bit toward the other way.
Poker players are pretty good at going with the flow. When you’re playing a poker tournament, everything is contingent because the tournament could last five more days or five more minutes. So, you’re like, “Yes, I’ll go to dinner with you, conditional on being knocked out of the tournament, and not entering this other tournament,” and x and y, and z. You’re very used to dealing with different stressors and contingencies and things like that. I think that’s somewhat unusual.
COWEN: If you’re very restless, does that seem to you like a bias in your decisions that you want to counteract with nudges, or something you should double down on? I would say you should double down on it, but what’s your view?
SILVER: I think closer to double down.
COWEN: Double down.
SILVER: Yes.
COWEN: So, what’s your main bias when you make decisions?
SILVER: I think I’m probably more emotional than people might think [laughs] I am on the surface —
COWEN: I would think you’re pretty emotional.
SILVER: Yes, that can come out on Twitter and things like that a little bit. People who are very competitive — and I put myself in that category, I think — can sometimes be a little bit irrational about continuing to fight even after they’ve already won. In some ways, I think that can be a bias potentially in certain ways.
COWEN: Who was the first human to think probabilistically with a reasonable degree of consistency? Or have we seen one yet?
SILVER: I think people are actually more intuitive, probabilistic reasoners than maybe the conventional wisdom holds. If you’re thinking about some hunter-gatherer — they’re reading context clues about where might be the particular wild beasts they want to kill and want to avoid being killed by. Gambling goes back very early in many cultures. Apparently not quite universal, but probably very common in most cultures.
COWEN: But that’s a sign they don’t think probabilistically because most gambling is negative sum, right? At best, zero sum, so we should see very little of it.
SILVER: It’s originally from divination, literally. Then it becomes gamified and commodified. It plays almost a spiritual role, I think, in ancient . . . I don’t know its anthropology super well. I’ve read a couple of books on it, but it has almost a spiritual divination role.
The notion of gambling halls or casinos, I think, dates back to maybe 16th- or 17th-century Europe, roughly. That’s potentially newer, but the notion of tempting one’s fate, which is still a romantic notion some gamblers have, like Erving Goffman, who I cite in the book. This is someone who has ennui or whatever. Maybe I’m using that term wrong, but they need to prove that they’re capable of taking a chance. It can be a little bit gendered. I think he thinks of it more as a post-World War II American male who feels like he has fewer chances to demonstrate his bravery. That’s always been one conception of gambling — a simulation of actual risk.
COWEN: Why shouldn’t people gamble only in the positive sum game? Take the US stock market — that certainly seems to be one of them — and manufacture all the suspense you want. Learn about the companies, the CEO. Get your thrill that way and don’t do any other gambling. Why isn’t that just better for everyone?
SILVER: Look, I’m not necessarily a fan of gambling for gambling’s sake. Twice a year, I’ll be in casinos and in Las Vegas a lot. Twice a year, I’ll have a friend who is like, “Let’s just go play blackjack for an hour and have a couple of free drinks,” and things like that. But I like to make bets where I think, at least in principle, I have an edge, or at least can fool myself into thinking I have an edge.
Sometimes, with the sports stuff, you probably know deep down you’re roughly break-even or something like that. You’re doing some smart things, like looking at five different sites and finding a line that’s best, which wipes out some but not all of the house edge. But no, I’m not a huge fan of slot machines, certainly. I think they are very gnarly and addictive in various ways.
COWEN: They limit your sports betting, don’t they?
SILVER: Yes, I’ve been limited by six or seven of the nine New York retail sites.
COWEN: What’s the potential edge they think you might have?
SILVER: It’s just that. If you’re betting $2,000 on the Wizards-Hornets game the moment the line comes out on DraftKings, you’re clearly not a recreational bettor. Just the hallmarks of trying to be a winning player, meaning betting lines early because the line’s early and you don’t have price discovery yet. The early lines are often very beatable. Betting on obscure stuff like “Will this player get X number of rebounds?” or things like that. If you have a knack for — if DraftKings has a line at -3.5 and it’s -4 elsewhere, then it can be called steam chasing, where you bet before a line moves in other places. If you have injury information . . .
It’s a very weird game. One thing I hope people are more aware of is that a lot of the sites — and some are better than others — but they really don’t want winning players. Their advertising has actually changed. It used to be, they would say for Daily Fantasy Sports, which was the predecessor, “Hey, you’re a smart guy” — the ads are very cynical — “You’re a smart guy in a cubicle. Why don’t you go do all your spreadsheet stuff and actually draft this team and make a lot of money, and literally, you’ll be sleeping with supermodels in two months. You win the million-dollar prize from DraftKings.”
Now, they don’t even try to advertise it as a skill game anymore, which in some ways is a mistake because the appeal to the male ego is a huge way to get new customers for sports betting and things like that. Every sports fan thinks he has some proprietary edge or knowledge or insight. +EV sports bettors are not a big lobbying group, or not a big cause, exactly.
I think there should be regulations on how much can you price-discriminate between your customers. Literally, they’ll take a million-dollar bet from somebody and wouldn’t take even a $50 bet from me on the same game, things like that, which also creates problems because money has a way to flow. Tyler, if you are able to bet a million dollars, and I’m able to bet $50, and I’m a winning bettor, then there are certain ways in which information will probably flow from me to you in an inefficient way, eventually.
COWEN: Sure.
SILVER: But no, the industry is very cynical in some ways. The old-school Vegas idea of “We’re going to post our line, and everybody gets a crack. You can bet X amount at a time. When you make a bet, then we may change the line and you can bet again” — That attitude has been lost. More of this European, UK style, where you’re doing lots and lots of customer segmentation and trying to maximize the amount of money you make from whales, which are bad degenerate gamblers, and not take much action at all from sharps.
Of course, you lose the price-discovery element of that. It’s a pretty interesting industry economically. What happens is, there are a few sharp sites that will say, “Hey, we will take $5,000 from anybody, and that helps us have better pricing and discover better lines.” Because by the time you get to kick-off or tip-off of a game, then you’re taking $50,000 bets or more. So, to have someone who’s willing to give someone a small +EV bet for a smaller amount because then you’ll get your retail customers later on — that was the old model.
COWEN: Shouldn’t they just stop the betting on, say, how many rebounds a player pulls down? It encourages corruption. It’s not actually a suspenseful act, whether it’s seven or eight rebounds, whereas who wins the Super Bowl presumably is a suspenseful, meaningful act for many people.
SILVER: That’s right. If you have a Jalen Brunson — I’m a Knicks fan — or someone who’s a major star, then it might make sense. But no, these books have way too wide a menu of events, and the sharper books don’t. The sharper books will pull a line down. If Steph Curry is questionable for the game, you can get a gigantic edge if you have any inside information about whether he’s going to play or not. They will just not take that bet, or they won’t do the more esoteric player prop stuff.
Or things like the NFL draft — if you are an NFL reporter for ESPN, then you could probably absolutely crush NFL draft bets. You might get fired if someone finds out, but the retail books are like, “Yes, we can put bets up that a knowledgeable person could make money on because we won’t let the knowledgeable people bet in the first place.”
Relatively, I guess it’s a sustainable equilibrium in the long run. I’m not quite so sure. One thing that sites like DraftKings are relying upon is, they’re basically piggybacking off the sharp sites, because you can just set your line based on Pinnacle, for example, or Circa. These are books that are sharp, and they actually do take bets from winning customers. DraftKings piggybacks off them.
COWEN: If we could enforce just an outright ban, what’s the cost-benefit analysis on banning all sports gambling?
SILVER: I’m more of a libertarian than a strict utilitarian, I think.
COWEN: Sure, but what’s the utilitarian price of being a libertarian?
SILVER: Look, I think most of the studies on gambling say that it’s a good transaction in terms of people getting some degree of enjoyment and excitement out of it, but maybe you have 5 percent of the population that becomes very addicted, and that 5 percent can account for a large share of volume. If it were me, if I were the gambling czar, I would probably ban slot machines and let everything else go, more or less.
The utilitarian part of me for slot machines is maybe enough to outweigh the libertarian part. They are deliberately designed to be addictive. It’s not very transparent what the odds actually are. You can’t actually go and it doesn’t say you have a -9 percent ROI at this particular slot machine. If you look at the demographics of slot machines, it’s much lower income. Sports betting and poker, which are the two forms I participate in myself, are mostly done by fairly high-income people who can, I think, afford to lose, but slot machines are nasty things.
COWEN: Speaking of scale, if we put you in a time machine, send you back to 1970, and you’re playing poker, why exactly is it that you would win so much?
SILVER: One way to put it is that poker actually has been solved more or less with game theory, with big, complicated programs called solvers that calculate the Nash equilibrium for any situation, given certain inputs.
COWEN: But when you’re at the table, that’s of limited use to you, right?
SILVER: No. If you think about the history of Texas hold ’em or poker, probably 99 percent of poker that’s ever been played, every hand that’s been played, counting online and things like that, has been played in the last 10 years. Poker’s not that old, the game. It goes back to the Mississippi riverboats. For many years, people weren’t really trying to use computers or anything to solve it. When Doyle Brunson played, he would literally, by hand, simulate hands by dealing out a deck hundreds of times to see, does a pair of twos beat ace-king off suit more often than not?
COWEN: Why isn’t there evolution across temperaments? Say people back then — they don’t know to be aggressive enough because they didn’t study game theory. But the ones who are aggressive enough naturally — they’re just going to win more money. It’ll become somewhat obvious over time that’s the way to play. You might be slightly better than the others in 1970 but not have a huge advantage. Is the abstract knowledge that powerful compared to market evolution?
SILVER: See, if I went back today —
COWEN: Today. Everything you know, time machine, you sit down, you’re going to do really well, right?
SILVER: — I would, I think, dominate the games. Yes, for sure. Anyone would.
COWEN: It’s because you know game theory and they don’t. Thomas Schelling was what, 1960; Von Neumann, Morgenstern just after World War II. Game theory is not that new. What exactly is it you know? To be a “Tight Aggressive” player with your strategy? None of them knew that? I feel I saw that in a lot of old movies.
SILVER: “Tight Aggressive” is part of it. It’s very hard to maybe intuitively know what the right bluffing frequency is in a certain spot. For many years, people would also give away lots and lots and lots of tells. If you go back and watch footage of the 1987 World Series of Poker, there are incredibly obvious tells. I think it’s just like there’s pressure on the market to compress it into a diamond or — what am I even talking about? I think we were just very early in the life cycle of the competitive pressures on the game in poker.
In some ways it’s a metaphor for other capitalistic ventures. I think you of all people may be underestimating the efficiency gains that are to be had when there are proper incentives and proper technology — basically, the combination and just the sheer volume of the number of nodes on the game tree that have been explored.
COWEN: Let’s say you’re a poker champion, and you want to quit. You can prove you are a champion. You’re still young. Who is it who wants to hire you in the actual world? Where would you go? A trading firm?
SILVER: Oh, hedge funds. Some of them, like Susquehanna or Jane Street in particular, are known for hiring ex–poker players.
COWEN: What’s the main flaw those individuals have as traders? Whatever their virtues may be. Chess players have flaws as traders, Math Olympiad types have flaws as traders. What are the flaws of the poker players?
SILVER: I have done a little bit of consulting for financial firms. My experience doing that is that it’s actually very similar because if you’re working for a trading desk, and you’re being asked, “What’s your opinion of how event X will affect opportunity Y to trade in the market?” It’s very much a game of estimation, and it’s very much a game of incomplete information, and it’s making decisions quickly, which is very poker-like.
There can be value in running a formal model and being more complete about something. That’s what I do for the election modeling-type stuff. But the trading opportunities pass very quickly, and the premium on speed relative to . . . journalism, actually, there’s a little bit of premium on speed too.
But to negatively stereotype, academia would be the contrast to that, where there, you can have the perfect answer, but it takes you a year to publish a paper, whereas can you have a good answer in five minutes? If you’re trading some commodity, and Joe Biden has a face-plant at this debate, you have to make a decision within a couple of minutes about that.
Likewise in sports betting, you see some line that clearly doesn’t make sense, and you have to infer, “Is this a golden opportunity? Or did Patrick Mahomes just get injured?” Or something like that. You have to make inferences very, very quickly. I think the poker players probably do pretty well in that environment.
COWEN: Now, you’ve run FiveThirtyEight. Now your newsletter has evolved into being a true business. Do you seek to hire top poker players, or you avoid them like the plague? Like, one is enough. “I’m enough. We don’t need any more of that skill. Get lost, fellow.”
SILVER: I would, for sure. Poker players are sometimes a little bit lacking in organization, and probably I’m a little bit lacking [laughs] in organization. Poker players are people who are sometimes a little bit irreverent, which is a quality I like.
The elections analyst I hired now — he’s great. Eli never failed to return a text message within 15 minutes or something like that. I don’t know when he sleeps. I need somebody like that because I’m running all over the place, and I can run a little hot and cold and be a little bit moody. I think you want someone who’s a little bit more steady for the newsletter.
I think also the writing skills are very important. I don’t know if this is becoming more important or less important in the world of ChatGPT, and so forth. Working on the book and then the newsletter, just to write every day, I feel like you can get 2x or 3x faster at writing, literally, if you just get a lot of reps in and things like that.
COWEN: Sure.
SILVER: It was the first time I’ve ever hired someone for this job, as opposed to at FiveThirtyEight. I was very excited by how many smart, mostly young people there were out there. Hopefully, I’ll find other ways to hire more of them in the future, but people who are a little bit more directed, I think, and organized.
COWEN: Running FiveThirtyEight and now your current ventures, what have you learned about social science, economics, game theory that you didn’t know to begin with?
SILVER: I feel like I’ve learned more in the past three years. Formally, I’ve been away from FiveThirtyEight for a year and change now. Informally, there was a lame-duck period where I was allowed, under my contract with Disney, to write a book. It’s probably focused —
COWEN: That’s this book.
SILVER: That’s this book.
COWEN: On the Edge, to repeat.
SILVER: Yes. It took three years. If you take three years and just talk to 200 really smart people, that’s an amazing thing to do. It’s what you do, but I wish I spent more of that time in my whole life. At FiveThirtyEight, the constant frustration was that we were brought in during a peacetime period for ESPN before we went to ABC News, where they felt like they had the world’s best business and could just print money, and we could be like a loss leader for them. So, you never actually had any incentives in place to run an actual business.
No one was going to lose their job because FiveThirtyEight lost X amount of money every year when you have a company that’s making hundreds of millions of dollars a year. Until you face headwinds, and cable TV and theme parks and all these things, all of a sudden, become very challenging businesses in a lot of ways. The movie industry during COVID and things like that. It’s a shame because, based on the early growth of the newsletter, I think FiveThirtyEight could have been a very good subscriber business.
No one was going to lose their job because FiveThirtyEight lost X amount of money every year when you have a company that’s making hundreds of millions of dollars a year. Until you face headwinds, and cable TV and theme parks and all these things, all of a sudden, become very challenging businesses in a lot of ways. The movie industry during COVID and things like that. It’s a shame because, based on the early growth of the newsletter, I think FiveThirtyEight could have been a very good subscriber business.
We tried to push that internally, but at a big company, the logic doesn’t make sense. They’re like, “Well, we have Hulu Plus launching, and therefore we can’t have another subscriber business that would distract from Hulu Plus.” It’s just very strange. I’m not sure how much of that is a result of being at Disney in particular, because they’re all about gigantic scale. It is theme parks. It is NBA and NFL contracts for hundreds of millions of dollars over a period of time. It’s being in every household in America, every movie theater.
I don’t think they saw the value of a small to medium-sized business that could still be a good business, a prestigious business, and profitable on its own merits, and that was frustrating. Given that I tend to be pretty incentive-driven, to have spent 10 years at a company — we basically had no incentive to work hard. Probably for the first eight years of that, I worked my ass off anyway.
But it’s just so nice to actually own your own work product. With subscriber newsletters in particular, the incentives are pretty linear. If you post more, you tend to get more signups in the long run, make more revenue. That’s a very nice change of pace and very motivating.
COWEN: Do you think that sports analytics has made sports too homogenized and more boring? Right now, for instance, the New York Knicks — they’re, in a way, trying to copy the Celtics. “Well, everyone should be a good shooter, and everyone should be a good defender.” In the 1980s or ’90s, teams were more different. Should I be happy about this?
SILVER: I think it’s the better version of the critique. I think the average NBA style is fairly attractive now.
COWEN: There are too many three-point shots, I would say.
SILVER: Maybe a little bit, but if you go watch footage of a game from the ’90s or something, this is more aesthetically appealing. It shows off some of the athleticism, I think, but the fact that you have less variety — this is true in poker too, by the way. There are fewer different varieties of loose-aggressive versus “Tight Aggressive” poker players. You can’t get away as much with having a flawed, eccentric strategy and then making up for it in other ways, which is a shame.
Maybe in the long run it means we’re not exploring different branches of the tree as well, and that can be problematic. But it’s a bad time — when things get hyper-efficient, then maybe the quirky players . . .
I saw a video yesterday about Major League Baseball batting stances, and there used to be guys that had very, very funky batting stances and things like that. Those have largely disappeared. If it cost you two-hundredths of a point of batting average — so you hit .289 instead of .291 — then there’s enough money on the line, and enough resources and training, and enough sports science where that’ll probably get beat out of you, potentially.
COWEN: Will Joel Embiid ever win a title?
SILVER: I know Daryl Morey pretty well. I think he’s probably a fan of yours. I’d say yes, I think.
COWEN: You think Paul George was a good idea? I think it was a bad idea. He’s 34. He’s had a history of being flaky. He’s a very talented player, but he doesn’t seem like the kind of guy to put you over the edge, and they need more of a glue guy.
SILVER: Daryl Morey is very good at finding the league-average players that can provide the glue guy. No, I think it’s a good move. Look, Daryl understands the importance of swinging for the fences. If healthy, that’s a very big three, and then he’ll find ways to acquire the minimum of salary guys or whatever loopholes there are on salary cap to get the $10-million-a-year guys. If healthy, that’s a team that can challenge the Celtics, I think. What’s the condition of probability in all three of their big three being healthy? Probably below 50 percent, frankly, but you don’t want the average outcome.
Also, the incentives in the NBA — and if you try to model out what the true discount rate is for sports general managers, they have some absurdly high discount rate, like 30 percent or something like that, or 50 percent, because their jobs last for five years on average.
COWEN: Yes, sure.
SILVER: So, I think it’s a fairly smart play to swing for the fences.
COWEN: Luka — overrated or underrated?
SILVER: I lost a fair bit of money trying to bet against the Mavericks in the playoffs, so I can’t call him overrated, I guess. I think Luka is properly rated.
COWEN: Properly rated?
SILVER: Yes.
COWEN: Who’s the most underrated player in the NBA, 2024 for our listeners?
SILVER: Let me think. I think Jalen Brunson in some ways. Now, that’s a Knicks fan coming out, although he was, I think, fifth or sixth in the MVP voting potentially this year.
COWEN: Yes, he was pretty high.
SILVER: As a Knicks fan, that’s one that comes to mind a little bit. Some of the Celtics — Derrick White, I think, is a very underrated player. I like Jayson Tatum and Jaylen Brown, but I think they are maybe respected as the 7th or 8th and the 23rd best player in the league, and they were a great historic team. I think the Holidays and the Derrick Whites and players like that are maybe top 30 NBA players — really, really good.
COWEN: I think I’d say Jokić, who, to me, is the greatest offensive player ever. Not necessarily recognized as such. Certainly, has a very high reputation, but maybe not as high as it should be.
SILVER: Even this year, you saw people were willing to turn on him a little bit. I think the consensus, now, of smart NBA nerds is that he’s the best player in the league. I think people might not realize that this is one of the highest peaks of any player in NBA history.
COWEN: Exactly, ever.
SILVER: I don’t think that’s widely recognized.
COWEN: How will AI improve sports analytics?
SILVER: I think it will improve things in areas like computer vision things and classifying plays and things like that. I think it will help a fair bit. Sports is weird because it’s in this medium data environment where it’s not quite a big data problem, although —
COWEN: But you could have more data if you put money into it, right?
SILVER: Yes. I think it’ll help more in the sports where they’re less linear. I think things like classifying strategies in soccer and football, where you have a lot of things happening at once, and you can’t have your nice little regression model, as compared to, say, baseball, where baseball’s very easy to represent with classical statistics. I think probably there’ll also be ways where it enhances the sports viewing experience in different ways. I think it’ll help sports video games a lot.
COWEN: Are hockey and combat sports going to use more analytics in the future?
SILVER: One of my friends, named Sunny Mehta, is an analyst for the Florida Panthers and assistant GM, and he’s an analytics guy, former poker player, former jazz musician, and just won the Stanley Cup in Florida. It’s transitioning in all the sports, I think.
Hockey probably, of the big four in the US, is considered the most old-school. Look, the market’s fairly efficient, and we’re long past the old boys’ network in most of the sports, probably in hockey. If you talked to Sunny, he’d say maybe two-thirds of the teams are fairly smart, and one-third of them are still dumb. I think we’ve probably passed the low-hanging fruit from Moneyball gains stage.
COWEN: Here’s a question I was asked during a Jane Street talk once. What is the most underrated basketball statistic for judging a player?
SILVER: It’s contextual because I think we’ve gotten to the point now where offensive rebounding has gotten underrated, actually. If you try to tease out the effect of rebounding in statistical models, then there are various things where different variables are correlated. Basically, I think if you look carefully, then offensive rebounding is actually quite valuable. You regain a whole possession. That’s worth a lot.
Whereas defensive rebounding — rarely is an individual player doing something right to get a defensive rebound. Whereas an offensive rebound requires a lot more skill and provides a lot more value because the expectation is 75 percent chance that you lose the ball. If you get an offensive rebound, then it’s like adding seven-tenths or eight-tenths of a point, basically. You get a whole possession back.
COWEN: My answer was team wins, at least if you’ve been on more than one team. Like, does a good team want you? That’s picking up your intangibles.
SILVER: Yes, uniquely in basketball, because it’s five players in a lineup at any given time, and a good player can control half the offense or more, literally — a Luka-type player or a Steph Curry-type player. I think it’s a more adequate test in basketball than other sports, or maybe for an NFL quarterback.
Baseball was my favorite sport for a long time, and there, it’s more individual, and there I was always fighting against the team-wins argument. I think it’s not a major indictment of Mike Trout that the Angels haven’t been very good, for example.
COWEN: But it’s not just you causing possible wins. Even before you’ve played for the team, the fact that they want you says something good about you. KCP has been wanted by good teams. His stats are good but not awesome, but that is a good sign about his quality.
SILVER: I think so, although we start to get into circularities when people become aware of that kind of thing. I feel this way a little bit with prediction markets, too. I consult now for Polymarket, so I’m officially a champion of prediction markets. When people believe, “Oh, the traders must be doing something really smart. Therefore, it’s trustworthy,” it can become circular logic sometimes, where you can make yourself vulnerable to thinking you have blind spots and having a bystander problem, basically.
Someone actually has to have alpha they’re providing, value-added they’re providing, and then other people can draft off that. But now and then, someone thinks that everyone else has vouched for somebody. It’s like the story of Sam Bankman-Fried in the book a little bit. Everyone assumed that because he has such prestigious coattails, that he’s hanging out with Anna Wintour and hanging out with Bill Clinton and Tony Blair, and he has his name up on a basketball arena in Miami, and he’s getting his praises sung by everybody in Sequoia Capital, and things like that.
At the heart of it, it was a big fraud. That’s a term I’m choosing intentionally.
COWEN: What’s your model of his risk-taking behavior?
SILVER: Like interviews with you. I think he is quite insane, irresponsible and insane, but I think he was very literal about being willing to take the St. Petersburg paradox bet, where if he could improve his version of utility by 2.00001x, he would literally say, “Hey, I’m a strict utilitarian. I’m willing to take that gamble.” In his interviews with me, he was almost proud of the high risk of ruin that he was taking. He would say things like, “If you’re not willing to actually, literally ruin yourself, where you’re literally a laughingstock, then you’re doing something wrong.”
COWEN: Is asking whether he might have been a sociopath a substitute for a risk analysis of his behavior, or is it a way of getting at the proper risk analysis of his behavior?
SILVER: No, I think that term is properly used because he also engaged with other people in running a giant company that people have deposited millions of dollars with and invested hundreds of millions of dollars in. I think that’s a proper frame. The book’s all about people who have this — I was going to say risk-taking gene. I realize it’s a little bit imprecise. I think it probably is genetic in part, but how is that paired with other different qualities? The most common pairing in the book is being really risk taking with really analytical, which is a very powerful, I think, pairing.
COWEN: You use this phrase — those people are in the river, on the river — what is it called?
SILVER: I called it the community of the river. The river is the virtual community of like-minded people, but that includes everything from poker to Silicon Valley. If you go further upstream, as I call it, even things like effective altruism and rationalism.
One thing that I found in the book is, I had this outline of “Here are all these areas that are interesting to me,” and literally started out . . . The first act I took for the book was, I flew to Florida — my first trip on a plane after the pandemic — and went to a giant casino in Hollywood, Florida, and played a poker tournament, and I branched out from there.
But the commonality you have between how all the different analytical nerds think, how the effective altruists and the hedge funders and the poker players and the VCs . . . It’s not exactly, exactly identical —
COWEN: Those are very different groups, though. Don’t some subsets of those groups get fooled to an especially high degree? Maybe not the hedge fund people — they face very strict discipline — but the effective altruists, the rationalists. There’s something to me a bit clueless about them, maybe systemically.
SILVER: Yes. The book has a very long discussion of EA, and in the spirit of EA, I am sure it’ll get reviewed on these different forums, and I hope it’s a fair —
COWEN: I mostly like EA, to be clear —
SILVER: I mostly like them too.
COWEN: — but there’s still some epistemic defect in the people that I observe regularly.
SILVER: I think they’re a little bit too trusting, is one thing. I think there is a lot to be said for having — skin in the game is a term that I didn’t invent. I think poker players — one of their best qualities is they have really good BS detectors. They’ve seen lots of smart SBF-like nerds who go on a winning streak, win a couple of tournaments, but are bullshitters and aren’t all they’re cracked up to be. I think EAs maybe lack that BS-detecting ability.
I think effective altruists are little bit too trusting, is one thing. I think there is a lot to be said for having — skin in the game is a term that I didn’t invent. I think poker players — one of their best qualities is they have really good BS detectors. They’ve seen lots of smart SBF-like nerds who go on a winning streak, win a couple of tournaments, but are bullshitters and aren’t all they’re cracked up to be. I think EAs maybe lack that BS-detecting ability.
COWEN: Do you ever see that old YouTube video when Ali G presents to Donald Trump — pre-presidential Trump — and Trump just dismisses him immediately and walks away? Trump has the BS detection ability. How does that fit with your model of the river people? Isn’t he, in other ways, almost the opposite of the analytical river people?
SILVER: I will say, if you take the EAs and put them in a separate camp, they are unusual because they have this analytical skillset, but they are less competitive. But then the movement gets usurped by this guy, Sam Bankman-Fried, who’s extremely competitive, and I think runs circles around them in some ways, even though he’s lying about things in other ways. Maybe competitiveness is good. The book is also against the Peter Singer form of selfless utilitarianism. I worry that people aren’t . . . Some of this comes from your critiques, I think, but people that aren’t a little bit partial — I worry that everything just falls apart on some level.
COWEN: Yes, there’s no incentive compatibility without some partiality.
SILVER: There’s no incentive compatibility. It’s interesting. I think Peter Singer, too, talks a little bit about how he almost thinks people that are somewhere on the spectrum are actually better moral reasoners in some ways.
COWEN: There’s evidence for this in some papers, in fact. They have less status quo bias. They’re more impartial in some ways.
SILVER: Yes. I think there might be something to that, and there are lots of people in the book that probably would be classified as being on the spectrum or would classify themselves that way.
Poker uniquely, I think, involves some combination of analytical skills and street smarts. If you totally lack the street smarts, then it’s still pretty hard to get ahead in a competitive world. You’re too likely to be taken advantage of.
COWEN: Now, your father was an academic. Did you get your street smarts from him or from somewhere else?
SILVER: Probably not. My dad is a sharp guy, and he would do things that were difficult problems. He was a Sovietologist back in the day and would try to interpret from rigged Soviet data what the real vital statistics were — how many births, deaths, abortions.
COWEN: So, he’s like you.
SILVER: Yes.
COWEN: Things have become more like criminology all the time in US politics.
SILVER: Yes. That involves making inferences from incomplete data. At one point he was offered to be part of a team to figure out how many deaths there had actually been in the Holocaust. Thankfully, he turned that down because that would be a thankless task for sure. He was starting out by working with data that was imperfect, inadequate, sometimes fabricated, and trying to find the real inferences from there.
A lot of it comes from being very competitive, being on a debate team and things like that in high school, playing poker for many years now, having a fair amount of life experience. We talked before about how I’m a little bit rambunctious and easily dissatisfied, and at this point, I think I’ve traveled — not like you, but been to a lot of different places and talked to a lot of people.
COWEN: What’s the stupidest risk you’re willing to take? It could be, “Oh, I’d go in a submersible with a billionaire.” Or “I’d take a helicopter ride every week.” What’s your stupidest risk?
SILVER: I’ve tried to become more healthy over time. This is a boring answer. I think I still probably have too much of a short-term bias in terms of not doing enough things to help extend my lifespan. I don’t know. One time when I was 10 or 11 years old at sleepaway camp, one of the instructors dared me to go walk around this pond with only a flashlight at midnight. For some reason, I did it. I don’t know why. I think they got fired. It was a big scandal or something.
COWEN: That doesn’t sound risky if you’re walking around the pond. You’re not walking on the winter ice, right?
SILVER: No. To be fair, you’re 11 years old and you’re by yourself and you have only a flashlight. It’s probably not the smartest thing in the world.
COWEN: I think my stupidest risks might just be road trips in places where the drivers are bad.
SILVER: Oh, for sure. Yes.
COWEN: I should do less of that.
SILVER: I went to Costa Rica a couple of years ago. It’s very rainy there, obviously, at least the part of Costa Rica that I was in. You’re probably doing things like going swimming. The waves in Costa Rica are really something. It’s probably, objectively, pretty dangerous to be in the Pacific Ocean without a lifeguard nearby. It’s probably fairly dangerous. Or going up some gravel road in the middle of a rainstorm. You wouldn’t want to take that risk constantly, but I think you have to live a little bit.
COWEN: What’s the way you would describe or think about what it adds to your life? I do know people who don’t do any of that. Maybe they’re missing something, but what is it exactly?
SILVER: I think probably it is genetic in part. There is some “thrill-seeking gene.” I like spicy food, [laughs] for example. I like travel. I like experiences that are more visceral. I think there has to be some innate component.
People I’ve talked to in the book — it’s mostly quants and things like that. It’s hedge fund people and poker players. There is one section of people who are physical risk-takers, people who are literally explorers. I talked to an astronaut, for example. If you talk to the people who were mountain climbing, scaling the seven summits, where you’re up 28,000 feet and you actually have some risk of dying — their community thinks there’s something genetic about it. They think it’s related to a certain type of stoicism, almost.
COWEN: What’s the richness they get added to their lives that maybe I would not? I’m somewhat risky with travel, but I’m not at all risky with physical activities.
SILVER: I think the notion of cheating death has to be thrilling. I think some of them are maybe a little bit depressed in some ways. It’s not quite self-destructive or suicidal but adjacent to that. Maybe you figure, “Well, I’m a little depressed, and so I can have a lot of fun and cheat death, or if I die, then maybe it’s not that bad.” I know it’s going to a dark place.
Some of Sam Bankman-Fried’s confidants, who are I think are in a very different category than these mountain climbers — they would say things like, “Well, he was miserable anyway, so he wouldn’t necessarily mind going to jail because in his day-to-day life, he was depressed and unhappy, so therefore . . .” That’s probably a pretty bad reason to be massively risk-taking, but I think there might be some element of that. It’s going in a slightly dark place.
But an Anthony Bourdain type — there’s someone who you would think of as being very risk-taking and probably liked traveling to places where it’s objectively pretty dangerous. I’m sure he had a very adequate crew with him and things like that.
COWEN: He never struck me as a risk taker. He was always going to places that look dangerous and aren’t, would be my sense of what he did.
SILVER: Maybe I’m revisiting now in light of his suicide. That’s a personality type. He liked to drink and do drugs and eat spicy food and things like that. He had, obviously, a penchant to travel. I think you see a cluster of personality traits there.
COWEN: Do you ever think much about how the broader world history of politics and political leaders may have been shaped by drugs and alcohol? There’s a book about how many of the Nazi commanders were high on different kinds of speed. I don’t know how true that is. There’s a book about the Soviet Union. It’s called Vodka Politics. Sam Bankman-Fried — there are various allegations about what he was taking. You see the photo of whatever on his desk. How important are drugs in all of this?
SILVER: I’m not a historian of drug culture or things like that.
COWEN: But you’ve talked to all these people. You’ve thought about it, right?
SILVER: Yes. I think, in the case of SBF, that drugs are maybe less a part of it than some things. I think in general that history is often quite contingent on very flawed human beings. There’s a good Matt Yglesias post about this in the context of decisions about whether people run for president or not, from Joe Biden deciding to step down to Teddy Roosevelt and things like that. When you do have some access to really smart people in the halls of power, people are often winging it. You know what I mean?
COWEN: Of course, usually.
SILVER: Yes.
COWEN: How can you not?
SILVER: Big decisions with limited information.
COWEN: And no second chances, really.
SILVER: Often, no second chances. I do think that people who are risk-taking tend to prevail for various reasons because they’re the people on the right tail of the graph. I think financial risk-taking probably has some positive correlation with risk-taking as far as substance use. Not that high, but maybe 0.2 or 0.3 or something like that. So, it wouldn’t surprise me if more important decisions than we think in the history of the world are made in various states of intoxication.
COWEN: If there’s status quo bias, should political leaders take drugs more?
SILVER: Before Joe Biden dropped out, it would’ve been a funny scandal. Maybe Joe Biden should have slept with a porn star and taken some shrooms. It would’ve changed his evaluation or the way people viewed him a fair amount. We’ve had a couple of teetotaling presidents, twice in a row, which is interesting.
COWEN: Trump is one of them. Yes.
SILVER: I’ll probably give the standard centrist left-of-center answer, which is, psychedelics are probably for some people pretty good, probably pretty dangerous for other people. Otherwise, I don’t know. I think some people are able to simulate different states of consciousness in different ways naturally. If drugs are one way to get to role-play a different person a little bit, then I think that probably is helpful for a fair number of people.
COWEN: Is the US presidency now in an age of what you might call parity in NBA terms? A series of one-term presidents, one after the other, just like the NBA champions keep on turning over and there’s no dynasty. We had this long era of dynasty — Bush, Clinton, my goodness, whatever else. Now it’s over and there aren’t going to be repeats. Or what’s that going to look like?
SILVER: I think part of it is a result . . . And yes, if Trump wins again, then he’ll be limited to one term. You would therefore have three one-term presidents in a row for the first time since, I think, 1892-ish. I was trying to look that up the other day. I think one thing that happened is that the canon slogan . . . I’m 46, so Carter was president when I was born.
The canon that you emerge into when you’re first covering politics is post–World War II politics, 1946 through Bill Clinton, 1996 or whenever. That was a period of historically very low polarization in the United States. You have lots of things going on. You have this de facto three-party system with Northern Democrats and Republicans and Southern Democrats, but you had relatively low partisanship, a lot of bipartisanship.
COWEN: Yes. A lot of things are going well.
SILVER: Things are going pretty well. It’s a good time for the US. Incumbent presidents get re-elected by landslide margins. Very often there’s a lot of goodwill.
If you go back, though, to the turn of the previous century — the Grover Cleveland years — every year, the election is separated by two points, and you bounce back and forth quite rapidly. That’s an era of high polarization and more like the era that we’re in right now. So, even before Biden’s age and other problems, the incumbency effect, if you tease it out carefully, has really dissipated.
In elections for Congress, we have a fairly large sample size. In the presidency, you may not have two incumbent presidents in a row who are voted out of office. Technically, it’ll probably be Kamala now, but yes, so there is precedent for it. This era of comity, this post–World War II era of comity that we all think of as being normal is actually more of the outlier in some ways.
COWEN: How long will it be before we have AIs who are better predictors than you are? Let’s say it’s a Centaur model. The AI sits down and says, “Nate, please talk to me for half an hour. Tell me your thoughts. Tell me what you know.” You give it your best half-an-hour worth of consulting on, say, the next presidential election, and then we just get rid of you. We send you off to pasture, and the AI does the work. You’re paid a small fee, and it does the actual prediction. How far away is that?
SILVER: I think election forecasting is an oddly hard problem, where the mathematical structure of it is fairly complex. The data is quite sparse, and you often have to make judgment calls based on incomplete information. That might be one of the relatively more robust areas.
COWEN: It gets half an hour of your time. It could pay for more if need be.
SILVER: Then where’s the line between me using an AI . . . In a very limited way, the book is AI-enhanced [laughs] in different ways, right?
COWEN: Yes.
SILVER: I am not convinced necessarily that we’re going to achieve superintelligence. I think it’s plausible-ish that it plateaus at very smart generalist human capabilities and is very good in certain areas, but I’ve become a little bit less convinced just in the last six months — since I had to send the book in to the printer, basically — of the inevitability of achieving superintelligence.
COWEN: Maybe it’s not super, but let’s say until an AI is as good as one of Phil Tetlock’s superforecasters. I would say that’s three to four years away, would be my guess. Closer to three, maybe even two. What would you say?
SILVER: I would take the over, I think, on that.
COWEN: Over as in 200? Or over as in, “Oh, it’s five years away”?
SILVER: 15 or 20.
COWEN: 15 or 20. That’s a long time. So, GPT-5 is not going to wow you?
SILVER: It wows me in certain way. As somebody who’s fascinated with language, I’m fascinated with the way languages are mathematized in different ways. In the book, there is a lot on AI, including reasons to be optimistic or pessimistic about P(doom). Your readers probably know what P(doom) is, I assume?
COWEN: Of course, yes. If they don’t know, they deserve P(doom), right?
SILVER: It does seem plausible to me that large language models reveal more about the mathematical structure of language than about just the intelligence of AI.
COWEN: They can now do Math Olympiad problems. There’s one forecast by someone working on this that, within a year, they’ll win Math Olympiads. It could be wrong, but when you say that it’s only a year away, I would take that pretty seriously.
SILVER: I think prediction is actually a pretty hard skill that probably life experience helps. Some types of prediction — if you have a phone book of millions of records, and you’re trying to predict what’s the likelihood that this spam email will be responded to and result in additional sale, I’m sure AI is already better than humans at that. But things that are less structured, I think, are harder for AIs and hard for humans.
COWEN: Why do businesses not seem to care that much about most predictions? Surely this has puzzled you. For 20 years, I’ve been waiting for prediction markets to take off. Obviously in politics, sports, it’s more than up and running, but in most areas, they’re pretty weak. What’s the actual problem on the demand side?
SILVER: I think this may be changing. One of the things I did in trying to figure out my next steps after FiveThirtyEight/Disney is, I talked to different financial firms and hedge funds and Wall Street firms about different types of consulting arrangements. Clearly, there are players in the market who are thinking carefully about political risk and other types of risks that you can price adequately by prediction markets.
I think there probably is some notion of critical mass when you have enough liquidity and enough volume to have relatively efficient pricing. I think we may now be at that threshold, but we weren’t four years ago or something like that.
COWEN: Take a truly important commercial question, like will China invade Taiwan, or when will they invade? I’ve looked at different markets or pretend markets on that. I just don’t take those seriously. It’s not that I feel I have a better estimate, but they don’t cause much updating for me.
SILVER: I think that the time horizons might be too long to be actionable necessarily. In the same way in sports betting, you can probably make money on futures bets. If I were to bet right now on who will win the NBA title in 2024 or ’25, there are probably some +EV bets there on the surface. However, that means I would have to tie up my capital for a year. There might also be some risk that the site I bet on goes belly-up or that I die or something. These long-term bets I think are —
COWEN: But that’s not very long term. That’s less than a year away. T-bill rates — they’re higher than they were, but they’re not that high. You have some marginal funds in things like T-bills. You’re in good enough health. You don’t want to do the submersible with the billionaire.
SILVER: But oftentimes, your edge in sports betting is two or three points, right?
COWEN: Yes.
SILVER: 2 percent or 3 percent as an ROI. If you can put it in the S&P 500 and make 8 percent, then it becomes a negative EV bet.
COWEN: But that gets back to my point. People should only bet, in essence, on US equities and nothing else. They’re fooling themselves in these other areas. If they want to spend their money that way, I’m libertarian enough, but I’m not that enthusiastic about it, and I don’t bet myself at all, really, on anything other than buy and hold and diversify.
SILVER: It is fun. [laughs]
COWEN: It’s not fun for me somehow. I’m a person who has never bet. I go to Las Vegas and I’m not even tempted. I don’t bet on any games. I don’t do slot machines. What’s wrong with me from your point of view?
SILVER: No, it’s probably rational in various robust and narrow senses of the term. Again, I’m not somebody who’s that big in casino gambling, but look, it’s the intensely competitive nature of things. There’s something called Enneagram. Do you know this, one of the different personality —
COWEN: Yes.
SILVER: The two rarest Enneagram types are the type that’s really analytical, basically, and the type that’s really competitive. If I take my Enneagram chart, those are two where I have the highest scores.
Typically, you think, “Oh, who’s someone who’s analytical?” Maybe they’re an actuary type. They’re calculating. Maybe they’re somewhat risk averse. Maybe they’re on the verge of being a little bit neurotic. That’s the stereotype.
The people in the book are not that. They’re analytical, but they want to . . . A lot of times they had things in their childhood. A lot of people — like Elon Musk or Jeff Bezos, for example — they had childhood trauma but not so traumatic, and they had enough privilege in different ways that they were able to still have people take their phone calls and things like that. But it’s often people where something, I think, went a little bit haywire, and they’re not necessarily strictly rational in some sense.
COWEN: Last two questions. First, where do you want to travel next?
SILVER: I have a good friend who’s Korean American. We’re hoping to go to Korea in December.
COWEN: That is cold. I’ve done that in fact. I love Korea, but I don’t know about December.
SILVER: I’ve been very little to the Middle East. I’d like to go there more. We were talking about Israel a little bit before.
COWEN: Last question. What is it you want to learn about next?
SILVER: One of the problems with writing a book where you take all the things that you really love is that you use a lot of your ideas. I’ll give you a dumb answer. I’d like to learn more about Pot-Limit Omaha, which is a form of poker that I haven’t played very much yet. I have this list of things. I want to learn more about European soccer, Pot Limit Omaha, and wine, basically. Those are all things that I think I would like very much, but I’ve never quite explored fully.
COWEN: Again, to repeat, Nate’s book, On the Edge: The Art of Risking Everything. Nate Silver, thank you very much.
SILVER: Thank you, Tyler.