Cognition9 min read

Decision-Making Under Uncertainty: Lessons from Poker Champions and Pilots

Goobeyond Research TeamApril 12, 2026

The best decision-makers are not luckier - they are better calibrated. Explore how pilots, poker players, and emergency room doctors make high-stakes decisions with incomplete information, and what you can steal from their playbook.

The Problem of Incomplete Information

Every important decision in your life is made with incomplete information. You cannot know the future. You cannot read minds. You cannot process all relevant variables. The question is not how to eliminate uncertainty - that is impossible - but how to make excellent decisions despite it.

Poker is the perfect training ground for this skill. Professional poker players make thousands of decisions per session, each with financial stakes, each with hidden information, each with time pressure. The best players are not luckier. They are better calibrated. They estimate probabilities more accurately, control emotions more effectively, and learn from outcomes more efficiently.

Annie Duke, a former professional poker player turned decision strategist, emphasizes that the goal is not to be right. The goal is to make the best possible decision given what you know at the time. Sometimes the best decision produces a bad outcome. Sometimes a terrible decision produces a good outcome. Learning to separate decision quality from outcome quality is the foundation of good thinking under uncertainty.

Mental Models from High-Stakes Professions

Pilots use structured decision frameworks to prevent cognitive overload in emergencies. The NAVigate model - Notice, Assess, Verify, Interrogate, Generate options, Act, Evaluate - forces sequential thinking when the brain wants to jump to action. The checklist culture of aviation exists because even experts skip steps under pressure.

Emergency room doctors practice what is called differential diagnosis: generating multiple possible explanations for symptoms before committing to one. This prevents premature closure - the cognitive trap of latching onto the first plausible explanation and ignoring contradictory evidence.

Poker players use expected value calculations to separate good decisions from bad ones regardless of outcome. If a decision has a positive expected value - meaning it wins money on average across many repetitions - it is correct even if it loses this particular time. This probabilistic thinking inoculates against outcome bias.

Building Your Decision Architecture

You can adopt these frameworks for everyday decisions. Start by keeping a decision journal: for significant decisions, record what you knew, what you believed, what you decided, and why. Later, review these entries to identify patterns in your thinking - confirmation bias, overconfidence, emotional interference.

Practice probabilistic thinking by assigning confidence intervals to your beliefs. Instead of 'I think this will work,' say 'I am 70% confident this will work.' This forces you to confront uncertainty explicitly and makes it easier to update your beliefs when new evidence arrives.

Build pre-mortems into your process. Before committing to a decision, imagine it is six months in the future and the decision has failed. What went wrong? This counteracts optimism bias and surfaces risks your excited mind would otherwise ignore.

Key Takeaways

  • Important decisions are always made with incomplete information
  • Separate decision quality from outcome quality to avoid outcome bias
  • Pilots, doctors, and poker players use structured frameworks to manage uncertainty
  • Decision journals, confidence intervals, and pre-mortems improve everyday thinking

Explore Related Assessments

Frequently Asked Questions

This is outcome bias - the tendency to judge decisions by their results rather than their process. The antidote is to evaluate the decision process independently: did you gather relevant information? Did you consider alternatives? Did you account for uncertainty? Did you manage your emotions? A good process can produce a bad outcome; a bad process can produce a good one. Over time, good processes win.

Overconfidence. Research consistently shows that humans systematically overestimate their knowledge, underestimate uncertainty, and overstate the precision of their predictions. The most dangerous decisions are often made with high confidence and low accuracy. Building the habit of expressing confidence as a probability range - 'I am 60-70% confident' rather than 'I am sure' - is a powerful corrective.

Emotions are information, not enemies. Anxiety may signal genuine risk. Excitement may signal opportunity. The problem is not having emotions; it is being controlled by them. The pause-and-label technique works: name the emotion explicitly ('I am feeling anxious about this choice'), take three slow breaths, then revisit the decision. Labeling emotions reduces their intensity and restores prefrontal cortex function.

Gut instinct is pattern recognition operating below consciousness. It is valuable in domains where you have extensive experience - an experienced firefighter sensing a building is about to collapse, a seasoned doctor recognizing an atypical presentation. It is dangerous in novel situations or domains where you lack expertise. Use intuition as one input among many, not as a substitute for analysis.