Annie Duke

Annie Duke is a former professional poker player and author who has become one of the most influential voices in the field of decision-making under uncertainty. She studied cognitive psychology at the University of Pennsylvania under Amos Tversky’s collaborators, left her PhD program to join her brother on the poker circuit, and spent twenty years as one of the top professional poker players in the world — winning the World Series of Poker Tournament of Champions and the NBC National Heads-Up Poker Championship. After retiring from poker, she translated her decision-making insights into consulting and writing, working with executives, athletes, and organizations on improving how they think about uncertainty.

Her first book, Thinking in Bets: Making Smarter Decisions When You Don’t Have All the Facts (2018), synthesizes poker strategy, cognitive science, and behavioral economics into a practical framework for better decision-making in any domain where luck and incomplete information play a role — which is to say, virtually all decisions that matter.

Thinking in Bets: Making Smarter Decisions When You Don’t Have All the Facts (2018)

The Core Diagnosis: Resulting

The book opens with a deceptively simple observation that has wide-ranging implications:

“Pete Carroll was a victim of our tendency to equate the quality of a decision with the quality of its outcome. Poker players have a word for this: ‘resulting.‘”

Resulting is the error of judging the quality of a decision by what happened, rather than by the quality of the decision process itself. The Pete Carroll example: Carroll’s call to pass rather than run at the goal line in Super Bowl XLIX is widely cited as one of the worst coaching decisions in NFL history. The pass was intercepted and the Seahawks lost. But the probability analysis of the specific situation, run by analysts, showed that the pass call was actually correct given the available information. The outcome was unlucky; the process was sound. The resulting error led an entire generation of sports commentators to mislearn a lesson about decision quality.

The opposite failure is also common: good outcomes produce false confidence in bad processes. Duke’s example is a CEO who drove drunk and made it home safely, and who updated his beliefs about drunk driving accordingly — toward greater tolerance for it.

“Drawing an overly tight relationship between results and decision quality affects our decisions every day, potentially with far-reaching, catastrophic consequences.”

See Resulting and Probabilistic Thinking for the full treatment.

The Two-Part Framework: Decisions and Luck

Duke’s foundational claim:

“Thinking in bets starts with recognizing that there are exactly two things that determine how our lives turn out: the quality of our decisions and luck. Learning to recognize the difference between the two is what thinking in bets is all about.”

Most outcomes in real life are a mixture of decision quality and luck. The problem is that we systematically attribute good outcomes to skill and bad outcomes to luck (for our own decisions), and systematically attribute good outcomes to luck and bad outcomes to poor judgment (for others’ decisions). These biases are mirror images of each other, and both prevent learning.

The corrective is to treat decisions explicitly as bets: choices made under uncertainty about which future will occur, with something of value at stake. Making this frame explicit helps in three ways:

1. It reminds you that any decision can produce a bad outcome even when well-made
2. It creates an epistemic norm for calibration: the person who wins bets over the long run is the one with more accurate beliefs
3. It triggers a more honest assessment of confidence levels before committing to a view

Beliefs as Bets

A key chapter shows how our beliefs are formed haphazardly and maintained through motivated reasoning. Rather than vetting information before believing it, we believe first and rarely revisit the belief:

“We hear something; We believe it to be true; Only sometimes, later, if we have the time or the inclination, we think about it and vet it, determining whether it is, in fact, true or false.”

The “Wanna bet?” trigger is Duke’s practical intervention: when you are asked to stake something of value on a belief, you suddenly treat that belief with more scrutiny. You ask: how confident am I really? What is the quality of my sources? What would I need to see to change my mind?

“Being asked if we are willing to bet money on it makes it much more likely that we will examine our information in a less biased way, be more honest with ourselves about how sure we are of our beliefs, and be more open to updating and calibrating our beliefs.”

The practical implication: expressing your confidence as a probability rather than as binary certainty (right/wrong) makes you a better communicator and a better learner. Saying “I’m 70% confident” rather than “I know this” creates openness to new information without requiring a humiliating reversal.

Truthseeking Groups

Duke argues that the individual effort to overcome resulting and motivated reasoning is insufficient. We need social structures that reward truthseeking:

“Recruiting help is key to creating faster and more robust change, strengthening and training our new truthseeking routines.”

Her model is drawn from her experience in poker: a small group of elite players who met regularly to analyze hands, share information, and rigorously evaluate each other’s decisions — without ego protection. The key features of a productive decision group (borrowing from Merton’s CUDOS norms for science):

• Communism: data belongs to the group, shared openly
• Universalism: evaluate information on its merits regardless of source
• Disinterestedness: vigilance against conflicts of interest, including emotional ones
• Organized Skepticism: structured encouragement of dissent

The social dynamic is crucial: the group rewards accuracy rather than ego confirmation, and members compete to be the best credit-givers and mistake-acknowledgers rather than the most confident. This reframes the social reward structure away from being-right and toward being-accurate.

Self-Serving Bias and Its Costs

Duke documents the predictable asymmetry in how we attribute outcomes:

“The way we field outcomes is predictably patterned: we take credit for the good stuff and blame the bad stuff on luck so it won’t be our fault.”

The cost is that this pattern prevents learning from experience. If every bad outcome is externalized as bad luck, the process that produced it is never examined. If every good outcome is internalized as evidence of skill, the role of luck is never accounted for — leaving us overconfident and unable to identify the factors that actually drove success.

The flip side: when evaluating others’ outcomes, the pattern reverses. Other people’s bad outcomes are their fault; their good outcomes are luck. The result is a systematic reduction of compassion and a distortion of learning from others.

Mental Time Travel: Backcasting and Premortems

One of the book’s most practical contributions is the operationalization of mental time travel — using imagined past and future selves to improve present decisions:

Backcasting: imagine you have already achieved your goal. What decisions and events got you there? This reveals what must be true for success and helps identify which steps are necessary versus merely assumed.

Premortem: imagine you have failed to reach your goal. What went wrong? This reveals potential failure modes that optimistic forward planning systematically misses. The premortem is the “organized skepticism” of project planning.

“Backcasting and premortems complement each other. Backcasting imagines a positive future; a premortem imagines a negative future. We can’t create a complete picture without representing both the positive space and the negative space.”

Ulysses contracts: named for Odysseus binding himself to the mast to resist the Sirens, these are precommitments that constrain future decisions. By binding future-self to present-self’s more rational choices, they create a decision-interrupt when in-the-moment irrationality would otherwise prevail.

The 10-10-10 Rule as Time Travel

Duke endorses Suzy Welch’s 10-10-10 tool as a practical implementation of mental time travel: for any decision, ask what the consequences will be in 10 minutes, 10 months, and 10 years. This interrupts temporal discounting — the tendency to over-weight immediate consequences relative to long-term ones — by forcing a multi-horizon evaluation.

Connections to Other Works

• Resulting and Probabilistic Thinking — the book’s foundational concepts in full
• Cognitive Biases and Heuristics — Kahneman’s work provides the scientific foundation for resulting and motivated reasoning
• Analogical Thinking and the Outside View — Epstein’s Range independently develops the outside view / reference class argument
• Measurement and Uncertainty Reduction — Hubbard’s framework for quantifying uncertainty is the measurement-science parallel to Duke’s probabilistic decision-making
• Rethinking and Intellectual Humility — Grant’s scientist mindset in Think Again is the epistemic parallel: both books argue for holding beliefs tentatively and updating readily
• David Epstein — Range and Thinking in Bets independently converge on the outside view / reference class as the corrective to inside-view overconfidence