4.3 KiB
aliases, Is A, Author, Topics, URL
| aliases | Is A | Author | Topics | URL | ||
|---|---|---|---|---|---|---|
|
Note | Annie Duke |
|
https://example.com/thinking-bets |
Thinking in Bets
Annie Duke
Duke, a former professional poker player turned decision scientist, builds her entire book around one insight: the quality of a decision and the quality of its outcome are different things, and confusing the two -- a mistake she calls "resulting" -- is the single most common error in human reasoning. A good decision made with the best available information can produce a bad outcome (because the world is uncertain), and a bad decision made recklessly can produce a good outcome (because sometimes you get lucky). If you judge decisions by their outcomes, you learn the wrong lessons: you abandon good strategies after a run of bad luck and double down on bad strategies after a run of good luck.
This framework is critical for anyone operating under uncertainty, which includes every founder, investor, and creator. When a newsletter edition underperforms, the instinct is to assume the content was bad. But maybe the content was excellent and the timing was wrong, or the subject line misfired, or there was a competing news event. Conversely, when an edition goes viral, the instinct is to assume you cracked the code. But maybe you just got lucky with the algorithm or the topic happened to catch a cultural moment. Duke argues that the disciplined response is to evaluate the decision process separately from the outcome and to learn from both.
Duke also introduces the concept of "thinking in bets" -- framing every belief and every decision as a bet with a probability attached. Instead of saying "this strategy will work," you say "I am 70% confident this strategy will work, and here is what I would need to see to change my mind." This probabilistic thinking reduces overconfidence, encourages intellectual humility, and makes it easier to update your beliefs when new evidence arrives. For someone making dozens of decisions a week about content, product, and business strategy, this is a powerful cognitive operating system.
Key takeaways
- "Resulting" -- judging decisions by outcomes -- is the most common reasoning error and leads to learning the wrong lessons
- Good decisions can have bad outcomes, and bad decisions can have good outcomes, because the world is inherently uncertain
- Think of every decision as a bet: what is your confidence level, and what evidence would change your mind?
- Probabilistic thinking reduces overconfidence and makes belief updating more natural and less ego-threatening
- Outcome fielding: after any result, deliberately separate the contribution of decision quality from the contribution of luck
- A "decision group" (trusted peers who challenge your reasoning) is one of the most effective tools for improving decision quality
- The goal is not to be right all the time but to be well-calibrated -- your 70% confident predictions should come true about 70% of the time
How I apply this
- After every major business decision (content pivot, pricing change, product launch), I do a decision journal entry that records: what I decided, why, what alternatives I considered, and what my confidence level was. When the outcome arrives, I compare it to the journal entry. This practice has helped me distinguish between decisions that were genuinely flawed and decisions that were sound but unlucky.
- I use probabilistic language in my strategic planning. Instead of "we will launch in Q2," I write "I am 80% confident we will launch in Q2, with the main risk being [X]." This forces me to think explicitly about uncertainty and plan for contingencies rather than operating on false certainty.
- When a newsletter edition performs unusually well or poorly, I resist the urge to immediately draw conclusions. Instead, I wait for a sample of 3-5 editions before changing my approach. This prevents me from overreacting to noise and ensures that any strategic change is based on a meaningful pattern rather than a single data point.