Notes on Thinking in Bets by Annie Duke

A while ago I had heard of the "poker book" Thinking in Bets: Making Smarter Decisions When You Don't Have All the Facts by Annie Duke, and I just recently finished reading it. It isn't really a book about poker; it's a book about decision science and how to make good/better decisions under uncertainty (i.e., pretty much all the time).

It was a nice summary of big lessons from a lot of other books like Thinking Fast and Slow that talk about the various types of cognitive mistakes we are susceptible to. I actually wished it had more juicy poker "guts" in it; while it had various stories and anecdotes, I wished there were more lessons drawn specifically from techniques or concepts poker players use or have developed. I would've gotten more out of this book compared to others like it if it had more of those "inner secrets" from poker; even without those, it was still an enjoyable and useful book.

My major notes and takeaways from the book are below.

---

​Intro: why this isn’t a poker book
Bet is decision about uncertain future
Outcome quality vs decision quality

1 life is poker not chess
Resulting: confusing quality of outcome for quality of decision
Hindsight bias
Reflexive vs deliberative mind
Von Neumann game theory 
I’m not sure is best response to many things

2 wanna bet
Make explicit that all decisions are bets
Betting against other future versions of ourselves 
The better/more accurate our beliefs, the better our bets
Hearing is believing by default
Preexisting beliefs change what we see
Being smart makes it worse
Being asked to bet on something makes you examine beliefs more and be more objective 
Express beliefs as probabilities 

3 bet to learn
Outcomes are feedback
Belief > bet > set of outcomes
Learning loop 
Separate luck from skill
Intermittent variable rewards manipulate
Self serving bias: blame losses on luck and wins on skill. Limits learning from experience
“If it weren’t for luck, I’d win every one”
“You can observe a lot by watching”. Don’t have to pay to learn yourself. 
Better to deconstruct own mistakes even when do well
reshaping habit
Cue then routine then reward
To break habit, keep cue and reward but insert new routine

4 buddy system
Good decision group, mentors
Truth seeking pod
Formal dissent channel

5 dissent to win
Merton
CUDOS
Communal ownership of data
Universalism: don’t judge message by messenger
Imagine getting same message from opposite type person
Or hide who messenger is
Disinterestedness 
Outcome blindness
Don’t tell someone how story ends
Reward people for disagreeing and argue other side
Skepticism
Anonymous dissent channel
Recruit devil's advocate
Express uncertainty when communicating 
I agree with you that in certain circumstances X (agreeing view) and … Y (opposing view)
Find out if people just want to vent or ask for advice
Focus on what to do better going forward

6 adventures in mental time travel
Imagine impact on future self
10/10/10
Consider impact on self in 10 minutes, 10 months, 10 years
Don’t look at life through in the moment zoom lens. Consider big picture more
The way we feel about outcomes is path dependent
Zoom lens of recent past distorts
Tilt in poker: when recent history affects mood and emotions take over good decision making
Precommitment contracts like Ulysses who asked to be tied to mast with wax in ears
Ulysses Contract
Decision swear jar
Illusions of certainty
Irrational resulting
Moaning about bad luck
Generalizations about personality
Echo chamber
“Wrong”
Scenario planning
Backcasting: working backwards from desired goals
Premortems: working backwards from negative future