Thinking in Bets - Annie Duke - Talks at Google
Uncertainty comes from...
- Hidden Information
- Chance
Our brain tend to make these mistakes...
- More prone to make false positive errors. We tend to see patterns everywhere - even if its not there.
- We work as tribes - we tend to make same decisions as the people we are surrounded by
- We don't try to falsify our beliefs - we have strong Confirmation Bias
Uncertainty in the data gives as a lot of room to interpret the data in a way that we want it to be(confirming our existing beliefs). Be wary of this.
Epistemic Closure: An understanding of what knowledge is true/trustworthy. We tend to get that from our tribe.
Epistemic Bubbles: We have access to only limited information. Echo Chambers: We have information - and people outside our group are not trusted sources.
One way to help you vet a bit of information that you distrust is by imagining it coming from a trusted source(or even without a source).
We should start Thinking in Bets - if we think of our belief/decision as bets, we would be more open to challenge/change it, rather than be tied to it.
Good way to challenge a belief is to ask "How much will you bet on that?"
Once we have a result, its very difficult for us to analyze the decision. The outcome will color the validity of the decision(Hindsight Bias). Try to analyze the decision/belief in spite of the result. "Did I make the right call with the information I had at that point?"
How to analyze decisions...
Ideally, do this in a group/with other people(a decision group). Other people can spot your bias way better that you can. Bias is a brain blind spot.
- Analyze the decision before there the outcome has occurred.
- When asking others to analyze the decision, only give them the information till the point of the decision. Don't tell them what happened after the decision.
- Try telling one person that there was a negative outcome and to another that there was a positive outcome. The analysis will be very different.
Our mind tries to find a coherent narrative(something that makes sense).
Not making a decision is deciding for the status quo. We tend to think of only new decision as actual decisions - but even not deciding is a decision.
Reasoning to be accurate: Lets create the most accurate version of the event Reasoning to be right: Lets create a version that makes me look right
Ulysses contract: Have a list of areas that you will generally be biased in(eg. I have a tendency to have a positive bias) - so that you can be more mindful in those areas - or even figure out ways to limit your choices in those areas.
People who are smarter are more susceptible to motivated reasoning. They can create really good stories with the data. Just because you know about the problem doesn't solve the problem. How often are you catching yourself making a mistake? How often are you making the mistakes?