Annie Duke – Wanna Bet? - [Invest Like the Best, EP.125]

At a Glance

Guest: Annie Duke, former professional poker player and expert in decision psychology. This episode’s core theme: how to break down decisions into five components—values, beliefs, decisions, randomness, and outcomes—and avoid judging decision quality by outcomes. Annie Duke argues that the biggest trap for decision-makers is “resulting”—using the final outcome to judge the correctness of the decision itself, while ignoring the interference of randomness.

The Foundation of Decisions: Values Define the “Rules of the Game”

Annie Duke believes that the starting point of all good decisions is clarifying personal values, because values determine what “game” you are playing. She points out that while everyone may share a common goal like “providing for the family,” the specific meaning of “providing” varies from person to person—some see it as money, others as quality time. Without clarifying this first, the subsequent decision-making framework cannot be established.

• Rule Setting vs. Goal Setting: Duke emphasizes that setting rules (e.g., “I am a long-term holder”) is more effective than setting goals (e.g., “I want to hold long-term”). Rules become internalized as identity, avoiding the cognitive burden of re-deciding every time a choice arises. For example, saying “I am a vegetarian” is easier to stick to than “I want to eat less meat,” because the former is a rule, while the latter is a goal requiring constant trade-offs.
• The Positive Power of Identity: Duke argues that tribe and identity are not solely negative. Citing research by J. Van Bavel, she notes that tribes provide a sense of “belonging” and “uniqueness.” In investing, “long-term holding” can be shaped as an identity label—“This is what we do, distinct from those who trade frequently.” This helps investors stick to discipline during market downturns, because breaking the rule means betraying their identity.

Anatomy of a Decision: Beliefs, Decisions, Luck, and Outcomes

Duke breaks down the decision-making process into five steps: Values → Beliefs → Decisions → Luck → Outcomes. She points out that people most often err between “luck” and “outcomes”—directly judging decision quality by the outcome while ignoring the mediating role of luck.

• Auditing Beliefs: Beliefs are the sum of “what you know” and “what you don’t know.” Duke emphasizes that people tend to overestimate the certainty of what they “know.” She suggests using the question “How do I know that?” for internal auditing, forcing oneself to consider the source and strength of evidence behind a belief.
• The “Resulting” Trap: Duke uses the example of Pete Carroll’s intercepted pass in the 2015 Super Bowl. The Seahawks chose to pass from the 1-yard line instead of handing off to star running back Marshawn Lynch, resulting in an interception. The public consensus labeled it the “dumbest decision ever.” But Duke points out that the probability of a pass being intercepted was only 1%-2%, while an incomplete pass would stop the clock and preserve two more offensive opportunities. “If you only focus on the outcome (the interception), you ignore the mathematical soundness of the decision itself—it was a good decision that encountered bad luck.”
• “Wanna Bet?”: This is Duke’s core thinking tool. When someone asserts a viewpoint, asking “Wanna bet?” forces them to consider: How certain am I? How much does luck play a role? What does the other person know that I don’t? This helps shift from an “inside view” (why I am right) to an “outside view” (why I might be wrong).

Group Decision-Making: How to Extract High-Quality Information from Others

Duke believes that individuals cannot overcome cognitive biases alone and must rely on others to calibrate their beliefs. However, group discussions naturally suffer from “belief contagion” and “outcome bias”—once you express your preference or know the outcome, others’ feedback becomes distorted.

• Isolating Outcomes from Beliefs: When reviewing a decision, Duke suggests “isolating the outcome”—do not reveal the final result when discussing decision quality. For example, lawyers should review strategies before the jury’s verdict, because the randomness of a single verdict is too high. Similarly, when seeking others’ opinions, do not state your own inclination first, to avoid triggering a tribal “agree” or “disagree” response.
• Communicating with Probabilistic Language: Duke recommends saying “I am 76% sure” instead of “I am sure.” This leaves room for others to offer differing opinions without triggering defensive arguments. She specifically emphasizes, never ask “Are you sure?” but rather “How sure are you?” —this shift in wording can significantly improve the quality of information extraction.
• Premortem: Before making a decision, assume the project has already failed, then work backward to identify the reasons for failure. This allows dissenting voices (especially minority opinions) to be incorporated into the decision-making process. Even if their views are ultimately not adopted, they feel heard.

Mentioned Positions

This section contains no substantive discussion of investable positions.

Judgments Worth Remembering

1. “Resulting” is the biggest enemy of decision quality (Annie Duke): Judging decisions by outcomes ignores randomness. Pete Carroll’s intercepted pass (1%-2% probability) was a good decision meeting bad luck, yet public opinion labeled it the “dumbest decision ever.”

2. Rules are more effective than goals (Annie Duke): “I am a long-term holder” is easier to stick to than “I want to hold long-term,” because rules become internalized as identity, avoiding the cognitive burden of re-deciding with every market fluctuation.

3. “Wanna bet?” is a powerful tool for calibrating beliefs (Annie Duke): When someone asserts a viewpoint, asking “Wanna bet?” forces them to consider certainty, luck, and counterparty information, shifting from an “inside view” to an “outside view.”

4. Isolate outcomes to get high-quality feedback (Annie Duke): When reviewing a decision, do not reveal the final result; otherwise, others will be influenced by “outcome bias” and cannot objectively assess the decision’s quality.

5. Replace “Are you sure?” with “How sure are you?” (Annie Duke): The former invites others to share uncertainty, while the latter triggers a defensive response. This shift in wording can significantly improve information extraction.

6. The most frustrating person may be the best teacher (Annie Duke): When facing an “irrational” opponent (e.g., someone playing chaotically at a poker table), instead of complaining, study their behavioral logic—this helps you identify patterns, adjust strategies, and improve all subsequent interpersonal relationships.

7. The goal of changing decision habits is to “make fewer mistakes,” not “make no mistakes” (Annie Duke): If someone used to panic 100% of the time when the market dropped 3%, and now does so only 80% of the time, that is a huge success. Small improvements compounded over the long term yield remarkable results.

8. “I am the fool”—admitting mistakes is a winner’s game (Annie Duke): Duke believes her best decision was “trying to find where each opponent was better than me.” This transformed her from an arrogant beginner into a continuous learner, and this mindset matters more than any specific technique.