Thinking in Bets: Making Smarter Decisions When You Don't Have All the Facts

Annie Duke, a former professional poker player with a background in cognitive psychology, draws on two decades of experience at the poker table and insights from behavioral science to argue that better decision-making begins with treating every decision as a bet on an uncertain future. Her central premise is that two things determine how our lives turn out: the quality of our decisions and luck. Learning to distinguish between them is the core skill the book aims to develop.

Duke opens by describing her path from academia to poker. She earned degrees from Columbia University and pursued doctoral work in cognitive psychology at the University of Pennsylvania. When illness interrupted her dissertation, she moved to Montana and began playing poker at the suggestion of her brother Howard, already a successful professional player. A temporary break became a twenty-year career in which she won a World Series of Poker gold bracelet, the Tournament of Champions, and the NBC National Heads-Up Championship, earning over $4 million. She frames poker as a laboratory for studying decision-making under uncertainty, where each hand involves up to twenty decisions and ends with a financial result that only loosely reflects decision quality because luck intervenes.

Duke introduces her argument through Seattle Seahawks coach Pete Carroll's controversial play-call in the closing seconds of Super Bowl XLIX in 2015. With the ball at the one-yard line, Carroll called for a pass instead of the expected handoff to running back Marshawn Lynch. New England intercepted, winning the game. Headlines condemned the call, yet analysts noted interceptions from the one-yard line occurred only about 2% of the time and the pass gave Seattle additional scoring opportunities. Duke argues Carroll made a high-quality decision that produced a bad result and introduces "resulting," a poker term for the error of equating decision quality with outcome quality. She pairs resulting with hindsight bias, the tendency to see known outcomes as having been inevitable.

She explains why these errors are pervasive by drawing on psychologist Daniel Kahneman's framework of System 1 (fast, automatic thinking) and System 2 (slow, deliberate thinking). Human brains evolved to create certainty and order, not to reason probabilistically, and our deliberative capacity is already overtaxed. The challenge is not to rewire our brains but to find practical work-arounds. Duke turns to mathematician John von Neumann, who modeled his foundational game theory on poker rather than chess. Chess has no hidden information; poker involves hidden cards, chance, and incomplete information. Life, Duke argues, resembles poker far more than chess.

Duke contends that embracing uncertainty is foundational. She defines a great decision not by its outcome but by a good process that accurately represents one's state of knowledge, and she proposes rating confidence on a zero-to-ten scale, which prevents the extremes of "right" and "wrong" and makes updating beliefs feel like small calibrations rather than wholesale reversals.

The book's second major thread addresses how beliefs drive decisions and how belief formation is flawed. Citing Harvard psychologist Daniel Gilbert's experiments, Duke explains that humans default to believing what they hear is true, even when information is explicitly labeled as false. This is an evolutionary artifact: Before language, beliefs came from direct sensory experience, and skepticism could be fatal. Once beliefs form, motivated reasoning entrenches them. Duke cites the 1954 study "They Saw a Game," in which Princeton and Dartmouth students watched the same film of a football game but reported dramatically different perceptions depending on their school loyalty. She notes that intelligence does not protect against bias and can worsen it, as more cognitively sophisticated subjects tend to exhibit larger blind spots in their own reasoning. As a corrective, Duke proposes the thought exercise "Wanna bet?" Even without literal wagers, this frame tempers certainty and prompts people to examine where their beliefs come from and how confident they really are.

Duke examines why learning from outcomes is difficult through the lens of self-serving bias, a concept from psychologist Fritz Heider: People take credit for good outcomes and blame bad ones on luck. Poker player Phil Hellmuth's statement, "If it weren't for luck, I'd win every one" (91), captures this tendency. Duke illustrates with Nick the Greek, a player she encountered early in her career who believed the worst starting hand was the best because of its surprise value. Nick occasionally won, reinforcing his flawed strategy, while attributing his losses to luck. He eventually stopped appearing at the game, and Duke learned he had been deported; she speculates his inability to learn from feedback may have contributed to his financial decline. Duke notes that we flip this pattern for others, as when Chicago Cubs fans singled out spectator Steve Bartman for sole blame after the 2003 playoff loss despite numerous on-field failures outside his control.

To combat self-serving bias, Duke profiles poker player Phil Ivey, who after winning a major tournament spent the celebratory dinner deconstructing every potential mistake rather than celebrating. Drawing on author Charles Duhigg's "golden rule of habit change," she argues we can keep the old cue (an outcome) and reward (feeling good about ourselves) but substitute truthseeking for credit-taking. She argues that a truthseeking group accelerates this shift. Drawing on the research of psychologists Philip Tetlock and Jennifer Lerner, she identifies three essential elements for a productive group charter: a focus on accuracy over confirmation, accountability with advance notice, and openness to diverse ideas. She stresses viewpoint diversity, citing legal scholar Cass Sunstein's research showing that ideologically diverse judicial panels produced significantly more moderate rulings. For rules of engagement, she adopts sociologist Robert K. Merton's CUDOS norms: shared data, merit-based evaluation, vigilance against conflicts of interest, and cooperative skepticism.

In later chapters, Duke introduces mental time travel as a decision-making strategy. She uses comedian Jerry Seinfeld's "Night Guy versus Morning Guy" bit to illustrate temporal discounting, the tendency to favor present gratification at future expense, and cites Stanford research showing that subjects who viewed age-progressed virtual images of themselves saved significantly more for retirement. Duke introduces the poker concept of "tilt," emotionally compromised decision-making triggered by recent outcomes, and presents Ulysses contracts as precommitment tools that bind present decisions to past rational intentions. Named for Odysseus, who ordered his crew to bind him to the mast to resist the Sirens, these contracts take practical forms such as automatic retirement deductions and preset loss limits. She also proposes a "decision swear jar" that treats words signaling irrationality, such as "I knew it" and "always," as prompts for reflection.

Duke closes by presenting three complementary planning methods: scenario planning (identifying possible outcomes and estimating their probabilities), backcasting (imagining success and working backward to identify how it happened), and premortems (imagining failure and working backward to identify causes). Drawing on psychologist Gabriele Oettingen's research, she notes that people who visualize obstacles are more likely to achieve their goals than those who engage only in positive thinking. She warns against hindsight bias using the metaphor of time as a tree: The trunk is the single past, and the branches are possible futures. As the future becomes the past, the present severs all branches that did not materialize, so the one outcome that occurred appears inevitable. She cites Judge Frank Easterbrook's opinion in a grain bin explosion case, in which he concluded the punitive damages assessed against the contractor West Side were "a consequence of hindsight bias." Duke argues that keeping an accurate record of what could have happened makes us better calibrators, more comfortable with uncertainty, and more content doing our best in a world that never guarantees favorable results.