Nate Silver on Risk-takers, Politicians, and Poker Players

Statistician Nate Silver, marking the release of On the Edge: The Art of Risking Everything, talks with Tyler Cowen about the community of analytical risk-takers he calls The River — poker players, sports bettors, hedge funders, effective altruists, and Sam Bankman-Fried among them — plus the economics of sports betting, why FiveThirtyEight never became a real business inside Disney, sports analytics and homogenisation, and how far AI still is from superforecasting.

Key ideas

1. A cross-cutting ‘community of the river’ unites analytical risk-takers — and explains how Sam Bankman-Fried fooled it. Silver’s term spans poker, Silicon Valley, hedge funds, and, further ‘upstream’, effective altruism and rationalism. He argues EA members are unusually trusting and lack poker players’ ‘BS-detecting ability’ — a product of having skin in the game — which let SBF’s prestigious coattails (Anna Wintour, Sequoia, an arena naming deal) substitute for real scrutiny.
2. Sportsbooks price-discriminate against winning bettors, sacrificing the price discovery that once made lines efficient. Silver has been limited by six or seven of New York’s nine retail sites for exhibiting the ‘hallmarks’ of a winning player — early bets, steam chasing, obscure prop bets. The industry has shifted from an old Vegas norm of one line for everyone toward European-style segmentation that maximises revenue from ‘whales’ while excluding sharps.
3. FiveThirtyEight never had real business incentives inside Disney. Brought in during ESPN’s ‘printing money’ era as a loss leader, it faced no consequence for losing money until Disney’s core businesses came under pressure — at which point internal logic (‘we can’t have another subscriber business, Hulu Plus is launching’) blocked a newsletter model Silver believes could have worked, in sharp contrast to the direct incentives of owning his own newsletter today.
4. Analytics has made sport more efficient and more homogeneous. Funky 1980s–90s NBA team identities and idiosyncratic MLB batting stances have been optimised away because the cost of an eccentric-but-suboptimal approach is no longer affordable — a dynamic Silver also traces in poker, where loose-aggressive eccentric styles are harder to sustain than they once were.
5. AI is a long way — 15 to 20 years, by his estimate here — from matching a top human superforecaster. Election forecasting is unusually hard for AI: sparse data and constant judgement calls under incomplete information. Silver has also grown somewhat less convinced of an inevitable path to superintelligence, suspecting large language models reveal more about the structure of language than about general intelligence.

Content

Simulating alternative lives and status quo bias

Cowen opens by turning Silver’s own book question back on him: across a thousand simulated versions of the world, how often does Nate Silver end up roughly where he is today? Silver estimates 20 per cent or less, tracing his actual path to a genuinely contingent event — the US government’s 2006 crackdown on online poker payment processing cost him his livelihood and pushed him, almost on a lark, to found FiveThirtyEight. He flips coins for trivial choices (dinner, at most a vacation destination) but stays deliberately analytical on career and life decisions, and argues his own restlessness is a trait worth doubling down on rather than correcting: poker tournaments train a specific comfort with total contingency, since every plan is conditional on when you get knocked out. Pressed on his own biases, he names over-competitiveness — a tendency to keep fighting even after already winning — as more of a live risk than status quo bias.

The economics and ethics of gambling

Silver draws a sharp line between gambling for its own sake, which he dislikes and would ban outright in the case of slot machines, and +EV betting where he believes, or can convince himself, he holds an edge. His utilitarian case against slot machines specifically: roughly 5 per cent of gamblers become addicted and account for a disproportionate share of volume, and the odds are deliberately opaque. He has been limited by six or seven of New York’s nine retail sportsbooks for exhibiting the hallmarks of a winning player — betting early before price discovery, ‘steam chasing’ a mispriced line, seeking out obscure prop bets. He describes an industry shift from an old Vegas norm — post one line, take any bet up to a cap — toward European-style customer segmentation that maximises revenue from ‘whales’ (degenerate gamblers) while excluding sharps, at the cost of the price-discovery role that once made betting markets more efficient; retail books like DraftKings now largely piggyback their pricing off sharp books such as Pinnacle and Circa. Cowen presses a stricter libertarian-adjacent line — bet only on positive-sum instruments like US equities and get your thrills there — and Silver concedes it is ‘probably rational’, while defending the intensely competitive appeal that keeps him in sports betting and poker specifically.

Poker’s professionalisation and the poker-to-finance pipeline

Poker has effectively been solved by game-theory ‘solvers’ that compute the Nash equilibrium for a given situation. Silver estimates 99 per cent of all poker hands ever played, across the game’s history, were played in the last ten years — pre-solver players like Doyle Brunson worked out odds by literally re-dealing a deck hundreds of times by hand. The gap between eras is not just abstract knowledge of game theory, which predates poker’s professionalisation by decades, but sustained competitive pressure exploring the full game tree — a dynamic Silver reads as a metaphor for efficiency gains under capitalism generally. That same analytical temperament makes ex-poker players attractive hires at trading firms like Susquehanna and Jane Street: both poker and trading reward fast decisions under incomplete information, in contrast to academia’s ‘perfect answer, one year to publish’ model. Poker players bring irreverence and street smarts but need organisational counterweights — Silver credits his own elections-analyst hire with a discipline (never missing a fifteen-minute reply window) that he says he lacks himself.

Running FiveThirtyEight and the incentive problem at Disney

FiveThirtyEight, Silver argues, never had real business incentives inside Disney and ABC. It was brought in during ESPN’s ‘printing money’ era as a loss leader, and no one lost their job over its losses until cable, theme parks, and COVID-hit film all became harder businesses at once. Even then, internal logic — ‘we have Hulu Plus launching, so we can’t have another subscriber business’ — blocked a newsletter model Silver believes could have worked, given the early growth of the paid Silver Bulletin newsletter after he left. He contrasts this directly with owning his own work product: with subscriber newsletters, incentives are close to linear — post more, get more signups, make more revenue — which he calls ‘very nice’ and ‘very motivating’ after ten years at a company with no such alignment.

Sports analytics, homogenisation, and underrated players

Asked whether analytics has made sport too homogeneous and boring — the Knicks trying to copy the Celtics’ shooting-and-defence model, rather than the stylistically distinct teams of the 1980s and 90s — Silver partly concedes the premise while defending the average product as more aesthetically attractive than the critique implies. He traces the same trend in poker, where loose-aggressive eccentric styles are harder to sustain profitably than they once were, and in baseball, where funky batting stances have largely disappeared once the cost of a suboptimal-but-distinctive approach (a few hundredths of a percentage point of batting average) became unaffordable given the resources now devoted to marginal gains. On specific players, he calls Jalen Brunson and Boston’s Derrick White underrated, agrees Nikola Jokić’s peak is under-recognised as one of the greatest in NBA history, and defends the Paul George–Sixers pairing as a rational ‘swing for the fences’ given how short general managers’ effective time horizons are. On analytics itself, offensive rebounding is underrated once its effect is properly isolated in a model — regaining a possession worth roughly 0.7–0.8 points, against defensive rebounding, which rarely reflects individual skill. He expects AI’s biggest near-term sports contribution to be computer vision and play classification, especially in less linear sports like soccer and football; hockey remains the most ‘old-school’ major US sport, with roughly two-thirds of teams now analytically sharp by his friend Sunny Mehta’s estimate.

Sam Bankman-Fried and the community of the river

The conversation’s centre of gravity is Silver’s account of Sam Bankman-Fried as a case study in On the Edge‘s central category: The River, his term for the cross-cutting virtual community of analytical risk-takers spanning poker, Silicon Valley, hedge funds, and, further ‘upstream’, effective altruism and rationalism. Silver reads SBF as having genuinely, not performatively, embraced strict expected-value utilitarianism to the point of courting a high risk of total ruin — SBF told Silver directly, in interviews for the book, that if you are not willing to actually ruin yourself, you are ‘doing something wrong’. Silver insists the word ‘sociopath’ is properly used here, not as a substitute for risk analysis but because SBF implicated other people’s deposited money in his own risk-taking. His diagnosis of effective altruism’s blind spot is a deficit of scepticism rather than of intelligence: EAs, he argues, are too trusting and lack poker players’ ‘BS-detecting ability’, a product of skin in the game and repeated exposure to plausible-seeming people who turn out to be bullshitters. SBF’s coattails — Anna Wintour, Bill Clinton, Tony Blair, arena naming rights, Sequoia’s endorsement — worked on EA and mainstream elites alike because social proof of this kind becomes circular: everyone assumes someone else has already vetted the person, so no one actually does. Both Silver and Cowen, who says he ‘mostly likes EA’, read the movement’s own utilitarian commitments as part of the problem: Silver’s book argues against Peter Singer’s fully selfless utilitarianism on the grounds that a system demanding total impartiality lacks incentive compatibility and ‘falls apart on some level’.

Drugs, contingency, and the shape of political history

Silver estimates a modest (0.2–0.3) positive correlation between financial risk-taking and risk-taking around substance use, and suspects more consequential historical decisions than commonly assumed have been made in some state of intoxication — while resisting the temptation to over-attribute SBF’s conduct to drugs specifically. He reads a run of one-term US presidencies (a Trump 2024 win would make three in a row, last seen around 1892) as reversion from an unusually low-polarisation, high-incumbency-advantage post-war era (1946–1996) back toward a more historically normal high-polarisation regime, in which the incumbency effect has ‘really dissipated’.

AI and the limits of near-term superforecasting

Asked how long before an AI outpredicts him, Silver holds election forecasting to be an unusually hard problem for AI — the data is sparse and success depends on judgement calls under incomplete information, unlike a well-represented problem such as spam classification. Pressed by Cowen for a concrete timeline to match a Phil Tetlock-calibre superforecaster, Silver answers 15 to 20 years, against Cowen’s guess of two to three. He has also grown somewhat less convinced, over the six months before this conversation, of an inevitable path to superintelligence, suspecting large language models may reveal more about the mathematical structure of language itself than about general intelligence — even as he grants current systems’ evident strength on structured, well-represented problems like Math Olympiad questions.

See also

• Nate Silver — speaker
• Tyler Cowen — host
• Nate Silver on Life's Mixed Strategies — Silver’s later CWT appearance, revisiting the river, AI timelines, and effective altruism
• On the Edge — the book this conversation marks the release of
• The River — Silver’s term for the community of analytical risk-takers, central to this conversation
• Effective Altruism — the movement Silver critiques as too trusting
• Risk Posture — decision-theory neighbour on calibrating risk appetite