Thinking in Bets by Annie Duke

Two things run your life

In Thinking in Bets, Annie Duke makes one claim and defends it for the whole book. Exactly two things decide how my life turns out: the quality of my decisions and luck. Learning to tell them apart is the entire skill.

Poker is where she learned it, and poker is a brutal teacher because luck is loud. You can play a hand perfectly and lose. You can play it badly and win. Chess hides this. Luck barely touches the board, so a loss usually points to a mistake. Life isn't chess. Most of my decisions involve hidden information and a lot of chance.

Duke calls the trap resulting: judging a decision by how it turned out. Its sidekick is hindsight bias, the urge to see what happened as inevitable. Every "I should have seen it coming" is hindsight bias talking.

The way out starts with three words that feel like weakness: I'm not sure. A great decision comes from a good process, and a good process starts with an honest picture of what I actually know. That picture is usually some version of "I'm not sure." So a decision isn't wrong just because it lost. If I weighed the odds in advance, one bad result means one possible future showed up. A sample size of one proves very little.

Everything is a bet, even the choices that feel too important to call bets. Where we live, which school, how we parent: each one wagers limited time, money, and attention on the future I want. And the opponent is rarely another person. It's the other futures I gave up when I picked one.

You can't audit yourself alone

Beliefs are bets too, and I place them carelessly. I hear something, I believe it, and only sometimes, much later, do I check it. Once a belief lodges, motivated reasoning guards it. I collect support and wave away the rest.

Being smart makes it worse. Blind-spot bias lets me see flawed reasoning in everyone but me, and it grows with intelligence. The better I am with numbers, the better I am at spinning them toward what I already think. That one stung.

The fix is a question from Duke's poker table: wanna bet? When someone challenges me to put money on a belief, I suddenly take inventory of my evidence. So I now say how sure I am, on a scale of zero to ten, or as a range of plausible outcomes. Almost nothing is 0 or 100. Admitting that is the first step to measuring uncertainty and shrinking it. It also lets me adjust a belief by a few points without feeling like I've flipped my whole identity.

Outcomes set the same trap. I credit my skill for wins and blame luck for losses, so I learn from neither. I flip the script for other people, which is also ego. Duke borrows Charles Duhigg's habit loop: keep the cue, keep the reward, swap the routine. The reward is feeling good next to my peers, so I change the scoreboard. I want to be the best credit-giver in the room and the fastest to admit a mistake, even when the result was good. Hardly anyone does that, which makes it feel exceptional.

I can't pull this off solo. A truthseeking group needs only three people: two to disagree and one to referee. It runs on a charter: accuracy over confirmation, accountability with advance notice, and openness to diverse ideas. Merton's CUDOS norms fill in the rest. Share all the data, judge an idea apart from its messenger, shield the group from the outcome and from my opinion, and treat skepticism as teamwork. The danger is a group of clones. Like-minded groups amplify bias, and the pull toward sameness is strong enough that I rarely notice it.

Decide with your future self in the room

The last big idea is mental time travel. In the moment, I'm impulsive. I'll take a small reward now over a bigger one later, a habit called temporal discounting. A bad result makes it worse. Poker players call it tilt: a recent event gets blown out of proportion and I react drastically.

The remedy is to interrupt the decision. Suzy Welch's 10-10-10 asks what each option looks like in ten minutes, ten months, and ten years. Flip it and ask how I'd feel today if I'd made the choice that long ago. Either way, past-me and future-me pull me out of the moment.

Then there's the Ulysses contract: bind future-me against my own impulses. Ulysses had his crew tie him to the mast. I can take a ride-share to the bar. I can also lower barriers, like carrying healthy snacks so the better choice takes less effort. Automatic retirement contributions are my favorite example. Set it once, and any change has to be a deliberate decision.

For big bets, scout the futures first. List the possible outcomes, with rough probabilities and how much I'd like each. Better still, backcast: imagine the headline "We Achieved Our Goal!" and work backward to how I got there. It's Ferriss' fear-setting and Covey's begin-with-the-end-in-mind in a new coat of paint.

Where I push back: Duke wants me to treat happiness like a long-term stock and ignore the daily ticks. I disagree. My happiness lives in the moment. I also think the Ulysses contract is a simple idea, think through the consequences, dressed up in jargon. And the later chapters drift away from betting.

Key takeaways

• Habits to build

  • Put a number on it. State confidence on a zero-to-ten scale, or as a range, before asserting a belief.

  • Treat my explanation of any outcome as a bet. Ask what luck did and what I controlled, on wins as well as losses.

  • Give credit out loud and admit mistakes first.

  • Lead with "and" instead of "but." "And" adds to what someone said. "But" cancels it.

  • Share more of the story than feels necessary when I ask for advice, and ask for reasons I might be wrong.

  • Express uncertainty so other people feel invited to disagree.

  • Treat the word "wrong" as a red flag. It's a conclusion, not a reason.

  • Ask "do you want to vent or do you want advice?" before helping.

• Exercises to try

  • Take one recent win and one recent loss. List what luck did and what I controlled in each.

  • Run 10-10-10 on a decision I'm facing this week.

  • Set up one Ulysses contract: automate a saving, remove a temptation, or make the better choice easier.

  • Write the "We Achieved Our Goal!" headline for a goal I care about, then backcast how I got there.

  • Map the possible futures of one big decision, with rough percentages.

  • Recruit two people for a truthseeking pod and agree on a charter.

• Quotes to live by

  • What makes a decision great is not that it has a great outcome.

  • We don’t win bets by being in love with our own ideas.

  • “Wrong” is a conclusion, not a rationale.

  • Acknowledging uncertainty is the first step in measuring and narrowing it.

How it connects to the Bible

Duke's favorite move is replacing certainty with a number. Fair. Almost everything I believe deserves one. The Bible claims one exception, and it's a person: “Jesus Christ is the same yesterday and today and forever” (Hebrews 13:8, NIV). Every other bet I place can lose. There's real delight in that. I can hold my plans loosely and still stand on something that doesn't move.