Lessons from Poker for Business Decisions

Annie Duke played professional poker for two decades before becoming a decision-making consultant. Her transition wasn't a career pivot -- it was a lateral move. Poker, she argues, is just applied decision science with faster feedback loops.

After reading her work and talking to several professional poker players, I'm convinced that poker contains more useful decision-making wisdom than any MBA program. Here are the lessons that transferred most directly to my work in engineering and business.

Lesson 1: Process Over Outcome

In poker, you can play a hand perfectly and still lose. You can play it terribly and still win. A single hand's outcome is heavily influenced by luck. Over thousands of hands, skill dominates -- but in any individual hand, the outcome tells you almost nothing about the quality of your decision.

This is poker's most profound lesson: judge decisions by the process, not the outcome.

In business, we do the opposite. A product launch that succeeds is celebrated regardless of the sloppy process that got there. A well-reasoned technical decision that leads to a bad outcome due to unforeseeable circumstances gets punished. We call this "resulting" -- working backward from the result to evaluate the decision.

Resulting is dangerous because it teaches the wrong lessons. If a reckless deployment goes fine, we learn "reckless deployments are okay." If a carefully considered architecture doesn't perform as expected, we learn "careful consideration is a waste of time."

The fix: separate your evaluation of the decision from the evaluation of the outcome. A good decision can have a bad outcome. A bad decision can have a good outcome. Judge the quality of the reasoning, not the quality of the result.

Lesson 2: Thinking in Probabilities

Poker players never say "I'll win this hand." They say "I'm a 65% favorite." The distinction is critical. Certainty doesn't exist; probability does.

In engineering meetings, I hear definitive statements constantly: "The migration will take two months." "Users will love this feature." "The system can handle the load." Each statement is implicitly claiming 100% confidence, which is never justified.

Better: "I'm 70% confident the migration takes two months. There's a 20% chance it takes three, and a 10% chance we discover a blocker that extends it to four."

This isn't just semantic. When you assign probabilities, you naturally consider alternatives. You start planning for the 30% case, not just the 70% case. You build contingencies. You make better decisions because you're engaging with reality rather than a fantasy of certainty.

Lesson 3: Pot Odds and Expected Value

Every poker decision comes down to pot odds: is the expected value of this action positive? If calling a $100 bet gives you a 40% chance of winning a $300 pot, the expected value is positive ($120 expected return vs. $100 cost). You call, even though you'll lose more often than you win.

In business, this translates to: not every positive-EV decision works out, but you should make every positive-EV decision anyway.

Example: Should you invest two weeks building a feature that has a 30% chance of increasing revenue by $50K/month?

Expected value: 0.3 x $50K = $15K/month of expected revenue for two weeks of engineering time. If the alternative is a feature with a 90% chance of $2K/month (EV = $1.8K/month), the riskier bet has better expected value.

Most organizations are too risk-averse. They consistently choose the high-probability, low-value option over the moderate-probability, high-value option. Poker teaches you to run the math and follow it.

Lesson 4: Tilt Management

"Tilt" is poker's term for emotional decision-making triggered by a bad outcome. You lose a big hand unfairly, and the frustration causes you to play recklessly on the next hand. The second loss is bigger than the first because it's driven by emotion, not analysis.

In business, tilt looks like:

• Overreacting to a customer complaint by derailing the roadmap
• Making a retaliatory hire after losing a key engineer
• Cutting corners on testing because the last release was late
• Abandoning a strategy after one quarter of poor results

Recognizing tilt is the first step. Professional poker players develop awareness of their emotional state and have rules for when to step away from the table. As an engineer, I have equivalent rules: don't make architectural decisions after an outage. Don't commit to deadlines when you're frustrated about a missed one. Don't respond to critical feedback within the first hour.

Lesson 5: Position Matters

In poker, acting last is a massive advantage because you have more information. You've seen what other players did before you have to decide.

In business, this translates to: be the last to speak in a meeting. Not because you're playing games, but because information accumulates as others share their perspectives. The person who speaks first has the least information. The person who speaks last has heard every argument.

This is especially important for leaders. When the boss speaks first, others anchor to that position. When the boss speaks last, they get honest input. Jeff Bezos reportedly insists on reading memos in silence before discussion starts, specifically to prevent anchoring.

Lesson 6: Bankroll Management

Even the best poker players go through losing streaks. Bankroll management ensures that a losing streak doesn't end your career. The rule: never risk more than a small percentage of your total bankroll on any single session.

Career equivalent: never put yourself in a position where a single failure is catastrophic.

• Don't take a job where you can't survive being laid off for six months
• Don't bet the company on a single product launch
• Don't stake your reputation on a single project
• Don't invest all your skill development in a single technology

Survivability first. Profitability second. This is the same lesson Buffett teaches about investing, and poker teaches it with faster, more visceral feedback.

Lesson 7: Table Selection

Professional poker players spend significant effort choosing the right table -- finding games where they have a skill advantage. A great player at a table of other great players will break even. The same player at a table of recreational players will print money.

Career equivalent: choose your environment carefully. Your absolute skill matters less than your relative advantage in a specific context. A strong generalist might struggle at a big tech company full of deep specialists. The same generalist might be incredibly valuable at a startup that needs versatility.

Pick tables where your specific skill set gives you an edge. Then play your game.

Putting It Together

Poker distills decades of decision-making into a game with rapid feedback loops. The lessons -- process over outcome, probabilistic thinking, expected value, emotional management, information advantage, survivability, and table selection -- apply directly to any domain involving decisions under uncertainty.

For a structured collection of decision-making scenarios that apply these same principles to real-world situations, the scenarios library on KeepRule organizes frameworks by decision type. It's a useful reference when you need to apply poker-style thinking to a specific business or career situation.

The cards are dealt. The question is whether you're playing with a strategy or just gambling.