Rules of Thumb — Thinking in Bets (Annie Duke)

Quick-scan, imperative heuristics distilled from the book. Each line carries a chapter citation. Scan the group you need; expand into practice via the implementation playbook.

Decisions & Outcomes

After any result, ask: "Was that a good decision or just a good outcome?" (Ch. 1)
Judge decisions by process quality, not by how they turned out (Ch. 1)
Treat every choice as a bet on an uncertain future — and remember not betting is itself a bet (Ch. 1)
Refuse "resulting": never let a bad outcome retroactively make a sound decision "wrong" (Ch. 1)
Refuse the reverse too: a good outcome does not make a bad decision "right" (Ch. 1)
Run the flip test — imagine the opposite outcome; if your praise/blame flips entirely, you were judging luck, not the decision (Ch. 1)
Model your situation as poker (hidden info + luck), not chess (complete info + skill) (Ch. 1)
When information was hidden, consciously resist anchoring on the outcome — that is when the reflex is strongest (Ch. 2)
Remember best decisions can have low success rates: picking the least-awful option among bad choices is still sound (Ch. 2)
Frame most decisions as bets against the alternative future selves you chose not to become (Ch. 1)

Beliefs & Calibration

Replace right/wrong with shades of grey — score decisions on a continuum, not a binary (Ch. 2)
Ask "how confident am I?" not "am I confident?" (Ch. 2)
Express beliefs on a 0–10 (0–100%) scale, or as a range of plausible alternatives (Ch. 2)
State the complement out loud: "I'm 60%" carries an explicit 40% chance of being wrong (Ch. 2)
Let range width signal information quality — tight = strong info, wide = weak info / high luck (Ch. 2)
Run the "Wanna bet?" challenge on your own strong claims to expose hidden uncertainty (Ch. 2)
Say "I'm not sure" — it is the accurate state under incomplete information, not a confession of ignorance (Ch. 1)
Update incrementally (e.g., 58% to 46%) so revising a belief never costs ego (Ch. 2)
Treat the only sin as failing to use contradicting evidence — not as being wrong in the first place (Ch. 2)
Hedge openly: calibrated uncertainty raises your credibility and invites others to add information (Ch. 2)
Hold even "facts" provisionally — many established facts later reverse (Ch. 2)
Watch for motivated reasoning: being smart amplifies bias rather than protecting against it (Ch. 2)

Learning & Fielding Outcomes

Field every outcome into a skill bucket or a luck bucket as the future unfolds (Ch. 3)
Watch for the patterned fielding error: crediting wins to skill, blaming losses on luck (Ch. 3)
Apply the same scrutiny to peers in reverse — and catch yourself doing it (Ch. 3)
Allow that any single outcome is part skill and part luck — almost never 100% either (Ch. 3)
Don't stop at the first flattering explanation; deliberately consider all plausible causes (Ch. 3)
Mine your winning outcomes for mistakes — good results can hide bad decisions (Ch. 3)
Remember misfielding is directional, so it compounds — but being predictable makes it correctable (Ch. 3)
Practice the hard habit on less painful material first (review wins before losses) (Ch. 5)

Working With Groups

Recruit a truthseeking pod — you cannot reliably de-bias yourself alone (Ch. 4)
Charter the group on three elements: accuracy over confirmation, accountability, and openness to diverse ideas (Ch. 5)
Treat ideological/intellectual diversity as load-bearing — disconfirmation is an emergent property of the group (Ch. 5)
Get consent before truth-telling — don't "Letterman" an unsolicited hypothesis on someone who didn't opt in (Ch. 3)
Keep a bet always looming: accountability flips disconfirming evidence from threat to useful (Ch. 5)
Adopt CUDOS — share all data, judge the idea not the messenger, guard against conflicts of interest, ask why it might be false (Ch. 5)
Run outcome-blind analysis: evaluate decisions while withholding the result so resulting can't contaminate the verdict (Ch. 5)
Engineer dissent — institutionalize devil's advocates and red teams rather than waiting for it (Ch. 5)
Lead with assent: open with genuine agreement and add with "and," not "but" (Ch. 5)
Let group approval install the habit, then internalize it so accuracy feels rewarding even when alone (Ch. 5)

In-the-Moment & Future Planning

Recruit your past and future selves into the decision via mental time travel (Ch. 6)
Run 10-10-10: weigh consequences in ten minutes, ten months, and ten years (Ch. 6)
Make the future-self vivid — a concrete future-self measurably changes present choices (Ch. 6)
Move regret in front of the decision, where it can still act as a brake (Ch. 6)
Use a pre-committed routine of standard questions, not in-the-moment judgment — you judge your impaired state poorly (Ch. 6)
Set Ulysses contracts: precommit decision-interrupts before you go on tilt (Ch. 6)
Keep a decision swear jar: catalog phrases that signal you're veering from truthseeking, so each one triggers reflection (Ch. 6)
Watch for tilt and ticker-watching — don't let recent fluctuations drive the decision (Ch. 6)
Scout the future: map a decision's branches, assign probabilities, and weigh probability against desirability (Ch. 6)
Backcast from imagined success — work backward from "We Achieved Our Goal!" to map the positive space (Ch. 6)
Run a premortem — imagine the goal failed and work backward to map the negative space (Ch. 6)
Work backward, not forward: standing at the end and "remembering" the path beats projecting forward (Ch. 6)
Find contentment in good bets, not good outcomes — stop being an outcome junkie (Ch. 6)

Key Quotes

"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." — Annie Duke, Introduction "We would be better served as communicators and decision-makers if we thought less about whether we are confident in our beliefs and more about how confident we are." — Annie Duke, Chapter 2 "Reconnaissance of the future dramatically improves decision quality and reduces reactivity to outcomes." — Annie Duke, Chapter 6

Related References

Implementation Playbook - how to sequence these in practice