#60 – Annie Duke: How to improve decision making, learning, & apply ‘backcasting’ for future success

As the skill gap between competitors narrows, luck has a larger proportional influence on outcomes because stochastic variability dominates single events, so only long series or marginal edges reveal true superiority.

Backcasting and pre-mortems make strategies more robust because working backward from success or imagined failure forces enumeration of counterfactuals and rare events, revealing hedges and contingency plans missed by forward-only forecasts.

Even with perfect information about probabilities, individual outcomes remain uncertain because stochastic variation controls single trials, so decisions must optimize expected frequencies rather than expect certainty on any one attempt.

Because the world is probabilistic, shifting toward probabilistic reasoning and catching overconfidence improves decisions by aligning choices with likely outcomes and preventing errors from deterministic assumptions.

When organizations punish unexpected bad outcomes more than they reward unexpected successes, people avoid non–status-quo choices because fear of blame pushes them toward low-volatility options that protect reputation and suppress innovation.

People commonly judge decisions by outcomes because it's cognitively simpler to infer quality from results than to reconstruct counterfactual probabilities, which causes luck to be misattributed as skill or blame.

Poker is a better model for real-world decision-making than chess because its incomplete information and role of luck mirror domains like medicine where decisions are made under uncertainty and uncontrollable events affect outcomes.

Quitting (folding) is an effective decision tool because exiting marginal positions limits further losses and preserves optionality for higher-expected-value opportunities, making it active risk management rather than surrender.