The Resulting Trap Escape
Judge decisions by process quality, never by outcome quality alone
The Resulting Trap Escape combats 'resulting'—the universal tendency to conflate decision quality with outcome quality. Annie Duke, drawing from professional poker, demonstrates that good decisions can produce bad outcomes and bad decisions can produce good outcomes because luck is always a factor. A poker player who goes all-in with pocket aces and loses to a lucky river card made the right decision—the outcome doesn't change that. Similarly, a drunk driver who gets home safely made a terrible decision despite the positive outcome. The framework teaches you to evaluate every decision based on the quality of the process—what information was available, how thoroughly it was analyzed, how well uncertainty was accounted for—rather than whether it happened to work out. This is counterintuitive because our brains are wired to learn from outcomes, and it requires building a new habit of asking 'Was this a good decision regardless of how it turned out?' Duke argues this single shift—separating decision quality from outcome quality—is the most important upgrade available to human decision-making.
- Decision quality and outcome quality are separate variables connected by probability, not certainty
- Excellent outcomes sometimes follow poor decisions, and poor outcomes sometimes follow excellent decisions
- Self-serving bias causes us to credit skill for successes and blame luck for failures, and reverse this pattern when judging others
- The quality of a decision can only be judged by the process used at the time it was made, with the information available at that time
- Reframe every decision as a bet with probabilitiesBefore making a decision, explicitly ask: 'How sure am I? What are the possible ways things could turn out? What decision has the highest odds of success?' This forces you to acknowledge uncertainty rather than pretending you know what will happen. Career changes, relationship commitments, and investments are all bets—choices made without complete information about future outcomes. Framing them as bets makes it natural to think probabilistically rather than in binary pass/fail terms.
- Record your decision process before you know the outcomeDocument what you knew, what you were uncertain about, what probabilities you assigned, and why you chose this option over alternatives. This creates an honest record that can't be corrupted by hindsight bias—the tendency to believe the singular outcome was inevitable when it was merely one of many possibilities. When you review the decision later, you evaluate the process, not the outcome. A journal entry that says 'I was 70% confident this would work because of X, Y, Z' is infinitely more useful than post-hoc rationalization.
- Conduct outcome autopsies that separate skill from luckWhen a decision produces a result, analyze how much of the outcome was due to controllable factors (skill, preparation, analysis) versus uncontrollable factors (others' actions, timing, random events, market conditions). Be especially skeptical of successes—self-serving bias makes you overattribute good outcomes to skill and bad outcomes to luck. For a fair assessment, ask: 'If I made this same decision 100 times, what range of outcomes would I expect?' If the actual outcome falls within the expected range, the decision quality was the constant, not the specific result.
- Build a truth-seeking group for honest decision evaluationIndividual bias correction is nearly impossible—blind-spot bias means you readily identify bias in others while remaining oblivious to your own. Duke recommends forming truth-seeking groups that follow Robert Merton's CUDOS framework: share data openly (Communism), apply universal evaluation standards, approach analysis disinterestedly, and practice organized skepticism. These groups hold each other accountable for resulting—catching when someone judges a decision by its outcome rather than its process.
- Evaluate decisions over a portfolio, not individuallyLike poker, life is 'one long game with many hands.' Any individual decision tells you almost nothing about your decision-making quality—the sample size is too small and luck variance too high. Instead, evaluate your decisions as a portfolio: over the last 50 decisions of this type, what was my hit rate? Over time, are my decision processes producing outcomes that trend upward? This long-term view dampens emotional reactions to individual wins and losses and reveals whether your process is actually working.
Chess involves minimal hidden information and luck—skilled players consistently win. Poker introduces incomplete information, random elements, and outcomes that depend on factors beyond individual control. A player with pocket aces—the best possible starting hand—can play perfectly and still lose to a lucky draw. Over thousands of hands, skill dominates. But any individual hand is a poor indicator of decision quality.
A person drives home drunk and arrives safely. By the logic of resulting, this was a good decision—the outcome was positive. But everyone recognizes this is absurd: the decision was terrible regardless of the outcome. The positive result was luck, not skill.
Duke describes two planning approaches: backcasting (imagining success and reverse-engineering the path) and premortems (imagining failure and identifying preventative factors). While backcasting is useful, premortems prove more effective because they promote realistic planning, proactive obstacle management, and psychological resilience toward setbacks.
Duke spent decades as a professional poker player, where the distinction between decision quality and outcome quality is visceral and constant. In poker, you can play the hand perfectly and still lose—and you can play terribly and win. Over thousands of hands, good decision-making produces good results, but any individual hand tells you almost nothing about decision quality. Duke observed that outside of poker, people almost universally judge decisions by their outcomes: a startup that succeeded must have been a good bet, a marriage that ended must have been a bad choice. This 'resulting' prevents accurate learning because it assigns credit and blame to the wrong factors. Her insight: 'If we aren't wrong just because things didn't work out, then we aren't right just because things worked out well.'