Thinking in Bets

Why Poker Is a Better Metaphor Than Chess

Most business books use chess as the metaphor for strategy: perfect information, deterministic outcomes, skill determines everything. Duke argues poker is the better metaphor because it involves incomplete information, probabilistic outcomes, and the constant interaction of skill and luck. Real life looks far more like poker than chess.

This distinction matters because it changes what good decision-making looks like. In chess, the right move is discoverable with enough analysis. In poker — and in life — the right decision is the one that has the best expected value given incomplete information. You can make the right decision and still lose. You can make the wrong decision and still win. The quality of the decision and the quality of the outcome are only loosely correlated.

Resulting Will Ruin Your Thinking

The book’s most powerful concept is resulting — the human tendency to judge decision quality based on outcomes. If a business decision works out, we call it brilliant. If it fails, we call it reckless. But the same decision made with the same information might produce different results depending on factors completely outside the decision-maker’s control.

Duke uses the example of Pete Carroll’s play call in Super Bowl XLIX. He called a pass play that was intercepted, and the entire sports world called it the worst play call in history. Duke argues it was actually a reasonable call given the game situation and probabilities. The outcome was bad, but the decision process was sound. The world confused a bad outcome with a bad decision.

This has enormous implications for organizational learning. If you reward or punish people based on outcomes rather than decision processes, you create incentives that actively damage decision quality. Employees learn to take safe bets that look good regardless of outcome rather than high-expected-value bets that carry risk.

How to Actually Think in Bets

Duke’s practical framework is deceptively simple: express your beliefs as percentages rather than as certainties. Instead of saying “this strategy will work,” say “I think there is a seventy percent chance this strategy will work.” Instead of “the market is going to crash,” say “I put a thirty-five percent probability on a significant market correction in the next year.”

This shift in language produces several benefits. First, it forces you to acknowledge uncertainty rather than pretending it does not exist. Second, it creates calibration. If you say you are ninety percent confident about something, you can track whether your ninety percent predictions come true ninety percent of the time. Most people are badly miscalibrated — their ninety percent predictions come true far less often than they expect.

Third, probabilistic thinking makes updating easier. If you hold a belief at seventy percent and encounter disconfirming evidence, you can move it to fifty percent without the psychological cost of admitting you were wrong. Binary thinking — right or wrong — makes updating feel like defeat. Probabilistic thinking makes it feel like calibration.

Outcome Fielding

Duke introduces a framework called outcome fielding, which asks you to sort every outcome into four categories: skill and good outcome, skill and bad outcome, luck and good outcome, luck and bad outcome. Most people instinctively attribute good outcomes to skill and bad outcomes to luck. This self-serving bias prevents learning.

The discipline of honestly assigning outcomes to skill or luck categories is uncomfortable but essential. That successful product launch might have been mostly lucky timing. That failed hiring decision might have been a reasonable bet that did not pay off. Without this honest sorting, you learn the wrong lessons from your experiences.

Decision Groups

The final major section of the book addresses how to implement these ideas in practice. Duke argues that individual decision-making is always compromised by motivated reasoning and confirmation bias. The solution is what she calls a decision group — a small group of people who have agreed to hold each other accountable for decision-making quality.

The key rules for effective decision groups: members must express beliefs as probabilities, they must share information before hearing the group’s opinion, they must actively seek disconfirming evidence, and they must be willing to be wrong. This is essentially a structured version of what good advisory boards and peer groups already do, but Duke provides specific protocols that make it actionable.

The Limitation

Duke is a better thinker than she is a writer. The poker analogies occasionally feel forced, and the book could be thirty percent shorter without losing content. Some of the decision-making research she cites — particularly the work by Kahneman and Tversky — is covered more thoroughly in other books. But her synthesis of poker thinking with behavioral science is original and useful.

Read This If…

You make high-stakes decisions under uncertainty and want a practical framework for separating decision quality from outcome quality. Especially valuable for managers, investors, and entrepreneurs.

Skip This If…

You have already read extensively on behavioral economics and cognitive biases. The core ideas will be familiar, though Duke’s poker-based framing adds a useful layer.

Start Here

Read the chapters on resulting and thinking in bets first. They contain the book’s two most actionable ideas. Then read the chapter on decision groups if you are in a position to implement one. The outcome fielding framework is worth practicing even if you do not read the full chapter.