THE RIGHT BET
CAN STILL LOSE

Most decision errors begin by collapsing every unknown into one vague feeling. Duke separates uncertainty into hidden information and luck. The first includes facts you do not yet have and may be able to investigate; the second remains variable even when the relevant facts are known. Before a consequential choice, make two columns. Put missing customer data, an opponent’s incentives, or an untested assumption in the first. Put market timing, another person’s final choice, or random variation in the second. Research what can be reduced, prepare for what cannot, and do not demand a prediction that the situation cannot honestly supply.

Imprecision is not the same as intellectual failure. When evidence is thin, Duke would rather see a broad probability range than a false point estimate—or the silent default of zero to one hundred. A range reveals what you know and creates a reason to learn. Begin wide, then earn precision. State the lower and upper bounds you can currently defend. List the evidence that would move either boundary. Research, ask others, and narrow only when new information deserves it. The value is not in pretending the first numbers are objective; it is in making uncertainty visible enough to update.

Echo chambers do more than hide information; they teach us which people may be trusted before a claim is examined. Duke invokes universalism as a corrective: separate the source from the truth value of the news. That does not mean ignoring expertise, incentives, or provenance. It means avoiding the shortcut in which identity decides the verdict. Run the source-swap test. Rewrite the claim without the speaker’s name, then imagine it delivered by someone you trust and someone you distrust. Ask what evidence would survive both versions. The exercise exposes whether your standard of proof changes with tribal comfort.

Duke calls it “resulting”: treating a good outcome as proof of a good decision and a bad outcome as proof of a bad one. Her football example changes emotional meaning the moment the imagined catch replaces the interception, although the choice made beforehand is identical. Grade the process with information available at decision time. Ask whether the options were identified, assumptions tested, probabilities considered, and risks proportionate. Outcomes still matter as data, especially over many comparable trials. One outcome, however, is too noisy to carry the entire verdict—and too powerful to be allowed to write the history alone.

Advice becomes contaminated when the adviser knows how the story ended. Once the result is revealed, people construct a causal path that makes it feel inevitable. Duke’s poker practice is to describe the hand only up to the decision point and stop. When seeking a review, give context, constraints, and the options you faced—but hide the choice and outcome. Ask the reviewer what they would do and why before revealing the rest. For a stronger test, give different reviewers opposite endings and compare their explanations. The gap between them measures how much the result, rather than the process, shaped the advice.

Group discussion can erase information before it combines it. In Duke’s hiring example, four interviewers should record their views separately before talking; otherwise the first confident opinion, especially a leader’s, becomes an anchor. Use independence before interaction. Ask each reviewer for a recommendation, confidence level, and decisive evidence in private. Reveal all assessments at once, then discuss where they diverge. The goal is not to prevent collaboration. It is to preserve the distinct observations that collaboration needs. Leaders should contribute last or submit their view in advance so their status does not silently choose the answer for everyone else.

Uncertainty gives the ego room to assign causes selectively. Duke captures the pattern sharply: a loss becomes bad luck; a win becomes genius. When judging a rival, the attribution often reverses. Audit explanations for symmetry. If luck can explain your loss, could it also explain part of your win? If skill explains a competitor’s failure, what skill may have contributed to their success? This does not require false modesty or denying competence. It requires applying the same causal standard across outcomes and people. A fairer story preserves the mistakes hidden inside success and the good choices hidden inside failure.

A group does not become truth-seeking merely because smart people meet. Duke extracts three operating rules from Seidel’s mentorship: a commitment to accuracy, mutual accountability, and openness to diverse viewpoints. Without explicit norms, belonging can reward agreement and punish useful dissent. Write the charter before the difficult decision arrives. Members may challenge reasoning without attacking identity; claims need evidence; confidence may change without loss of status; leaders accept correction; disagreement is recorded rather than smoothed away. The charter matters only if the group can point to a moment when it changed a conclusion or exposed a comfortable story.

Success is the hardest outcome to interrogate because it has already paid the ego. Duke proposes a team ritual: present the week’s best success and identify five things done wrong along the way. The number is a prompt, not a magic quota. Run a success post-mortem before celebration closes the case. What worked by design? What worked despite the process? Which risk failed to materialize? What would break at larger scale? Preserving those distinctions prevents luck from being promoted into strategy, keeps weak steps visible beneath a strong result, and makes the next win more repeatable.

Knowing a bias will appear is not the same as resisting it in the moment. Duke uses the Ulysses contract: a choice made by the calmer present self that constrains a predictable future weakness. Her examples include booking exercise classes, controlling what food enters the house, or removing the option to drive after drinking. Place the guardrail before the temptation. Pre-commit a deadline, automate a transfer, ask a colleague to hold the launch criteria, or make the risky option unavailable. The contract should protect a considered intention, not imprison you against genuinely new evidence; define in advance how it may be revised.

Duke does not ask people to deliberate over every move. Fast judgment is necessary, and experienced intuition may be valuable. Her condition is accountability: at some point, the gut must be able to explain itself. After an intuitive choice, reconstruct the cues, patterns, and trade-offs well enough to teach them to another person. If the explanation ends at “it felt right,” mark the judgment as a hypothesis rather than expertise. Look for the signal the intuition may have compressed and the bias it may have concealed. Periodic explanation turns tacit knowledge into something a team can test, transfer, and improve.

Learning the names of biases does not remove them. Duke openly says she still misses them; progress is catching the pattern more often and a little faster. That standard is both stricter and kinder than pretending the mind has been fixed. Measure recovery, not purity. After a defensive reaction, a result-driven story, or an overconfident forecast, note when you noticed it and what interrupted it. Reward the catch, revise the decision if possible, and make the cue easier to recognize next time. Small improvements in repeated choices can compound, while shame often creates another reason to hide the evidence.