How to Make Better Decisions: The Decision Science Framework That Reduces Regret

By Gwyndalyn Henderson

The finding that behavioral economists most consistently fail to communicate is this: the cognitive errors that produce the worst decisions are not made by uninformed, careless, or irrational people. They are made by smart, experienced, well-intentioned people whose thinking is precisely the kind that the situation appears to call for. Daniel Kahneman at Princeton, Gary Klein in applied decision research, Paul Slovic at the University of Oregon, and Annie Duke in professional decision-making all converge on the same uncomfortable point: the decisions you feel most confident about — the ones that feel obviously right, where the evidence seems clear and the path forward feels certain — are exactly the decisions where the most costly biases are most active. Confident judgment is not a reliable signal of accurate judgment. The goal of making better decisions is not to trust your instincts more. It is to build the specific structural practices that catch the systematic errors your intuition is most likely to miss.

If you want the complete decision frameworks and mental clarity system built around these principles, The Focused Mind applies them to the financial, professional, and relational decisions where clear thinking matters most.

Kahneman: System 1, System 2, and Where the Errors Live

Daniel Kahneman, Nobel laureate in economics and professor emeritus at Princeton University, spent decades documenting the systematic patterns in human judgment and decision-making. His framework, popularized in Thinking, Fast and Slow, distinguishes two modes of cognitive processing: System 1 (fast, automatic, intuitive, associative, emotional) and System 2 (slow, deliberate, effortful, analytical). Most decisions — including most important ones — are made by System 1 with post-hoc rationale supplied by System 2. The errors in judgment that Kahneman's research catalogs are almost entirely System 1 errors: pattern-matching applied in situations that require analysis, emotional responses substituted for probability estimates, representativeness used in place of base rates.

The most important of Kahneman's findings for practical decision improvement is not the existence of biases — it is their resistance to awareness. People who have been told about confirmation bias continue to exhibit it immediately afterward. People who understand anchoring are still anchored by irrelevant numbers. Knowing that a bias exists does not significantly reduce its effect on your behavior, because the bias operates in System 1 before the System 2 knowledge is engaged. The implication is that debiasing through awareness — the standard approach of most "think better" advice — is largely ineffective. What works is structural: building decision processes that force the right questions to be asked before the System 1 judgment is finalized, regardless of whether you feel the need to ask them in the moment.

The scroll-stopper from Kahneman's research: in one of his most replicated findings, experienced financial analysts showed the same anchoring bias as novices when asked to estimate security values after being shown arbitrary random numbers. Expert experience does not immunize against System 1 errors; it sometimes amplifies them by building the fluency that signals false confidence.

Klein: When Intuition Is and Isn't Reliable

Gary Klein, a cognitive psychologist who has spent his career studying expert decision-making in high-stakes real-world environments — military commanders, firefighters, intensive care nurses, chess grandmasters — provides an important and often missing counterweight to the Kahneman bias-focused literature. Klein's research on naturalistic decision-making shows that expert intuition is genuinely reliable in specific conditions, and that blanket distrust of intuitive judgment is as costly as blanket trust in it.

Klein's research identifies the conditions under which expert intuition can be trusted:

• High-validity environment: the domain has stable patterns that reliably predict outcomes (chess positions predict likely moves; fire behavior follows physical laws)
• Adequate experience: the decision-maker has had substantial experience with similar situations and received clear feedback about whether their judgments were correct
• Rapid feedback loops: mistakes produce consequences that are visible and attributable quickly enough to update the intuitive model

The conditions under which expert intuition is unreliable are equally specific: low-validity environments (stock prices, social dynamics, long-term career predictions), delayed or ambiguous feedback (where the quality of a decision is not visible for months or years), and situations where outcomes are affected by factors outside the decision-maker's control. Kahneman and Klein collaborated on a paper reconciling their findings — and agreed that the critical question is not "should I trust my intuition?" but "what kind of environment am I in, and does it meet the conditions for expert intuition to be valid?" Overriding expert intuition in a high-validity environment is costly. Trusting intuition in a low-validity environment is equally costly. The decision-improvement task is accurate diagnosis of which situation you are in.

Slovic: The Affect Heuristic and Emotional Decision-Making

Paul Slovic, professor of psychology at the University of Oregon and president of Decision Research, has produced extensive research on the affect heuristic — the cognitive shortcut in which people substitute an emotional response for an analytical judgment. When asked to evaluate the risk or benefit of an activity, technology, or choice, people tend to answer the question "how do I feel about this?" and report the answer as if they had answered "what does the evidence show about risk and benefit?"

Slovic's research documents a consistent inverse relationship: when people feel positively about something, they simultaneously estimate its risks as lower and its benefits as higher, even when the actual risk and benefit information is unchanged. When the same technology is described with emotionally positive framing, its risks are estimated as lower than when it is described with neutral framing. The affect heuristic is not a failure of rationality in emotional people; it operates reliably in experts, in people who believe they are being analytical, and in contexts where the person is actively trying to be objective. It is a feature of how the brain processes information, not a flaw in specific individuals.

The implication for decision improvement is specific: any decision in which you have a strong pre-existing emotional response — a person you like or dislike, a plan that feels exciting or threatening, an investment that feels safe or risky — requires deliberate separation of the emotional response from the evidence evaluation. The emotion is information about your values and preferences; it is not information about the actual risk-benefit analysis of the option.

Duke: Resulting and the Process-Outcome Distinction

Annie Duke, former professional poker player and author of Thinking in Bets, brings the decision theory of professional gambling to the question of how people evaluate their own decision quality — and identifies one of the most costly and least recognized evaluation errors: "resulting," the tendency to judge the quality of a decision by its outcome rather than by the quality of the process that produced it.

Resulting is costly in both directions. A good decision that produces a bad outcome due to bad luck is evaluated as a bad decision — and the person abandons or modifies a sound process. A bad decision that produces a good outcome due to good luck is evaluated as a good decision — and the person doubles down on a flawed process. In domains where luck and variance are substantial (financial decisions, career decisions, relationship decisions, health decisions), outcome-based evaluation of decision quality is systematically misleading. It trains the decision-maker to optimize for what happened to work last time rather than for what maximally increases the probability of good outcomes on average. Duke's research on poker decision-making, and her work applying it to everyday decisions, consistently shows that improving decision quality requires separating process evaluation ("given what I knew at the time, was this the best decision I could have made?") from outcome evaluation ("did it work?"). Both questions matter — but conflating them produces systematic deterioration in decision quality over time.

Strategy 1 — The Pre-Mortem

Psychological mechanism: Prospective Hindsight (Gary Klein's research — imagining that a decision has already failed and working backward to explain why activates causal reasoning that forward-looking planning suppresses; it reliably surfaces risks that optimism bias and confirmation bias keep invisible during standard decision analysis). Kahneman, who has described the pre-mortem as one of the most effective debiasing tools he knows, notes that it works specifically because it gives permission for dissent in a group setting and because the prospective hindsight frame activates different cognitive resources than forward-looking analysis.

Implementation steps:

• After reaching a decision but before committing to it, run a 10-minute pre-mortem: "It is one year from now. We made this decision. It has failed. What went wrong?"
• Write every failure mode you can identify — do not filter for plausibility; the goal is to surface possibilities that optimism is suppressing
• For each failure mode: is this preventable? If yes, what is the specific mitigation? If no, is the downside acceptable?
• Adjust the decision based on the pre-mortem findings — either by modifying the plan to reduce the risk or by explicitly accepting the identified failure modes as acceptable given the expected benefit

Strategy 2 — The Decision Journal

Psychological mechanism: Feedback Loop Construction (Duke — resulting can only be corrected by explicitly recording the decision process at the time of decision and comparing it to the outcome later; this builds accurate calibration between confidence and accuracy over time). A decision journal is not a diary or a to-do list. It is a structured record of decision quality that provides the feedback loop that real-world decision-making often does not naturally supply — because outcomes are delayed, ambiguous, or affected by factors that obscure the relationship between decision quality and result.

Implementation steps:

• For significant decisions, record before acting: the decision, the alternatives considered and why they were rejected, the key assumptions the decision rests on, your confidence level (1-10), and what specific evidence would tell you the decision was wrong
• Set a calendar reminder to review the entry at the relevant outcome point — 30 days, 90 days, or 1 year, depending on the decision time horizon
• At review: what actually happened? Which assumptions were correct? Where was your confidence calibrated correctly, and where were you overconfident?
• Over time, look for patterns in your miscalibration — the specific types of decisions where you are consistently overconfident or underconfident — and apply additional scrutiny to future decisions in those categories

Strategy 3 — The 10/10/10 Temporal Reframe

Psychological mechanism: Temporal Discounting Correction (Kahneman — present emotional state dominates decision-making in ways that systematically underweight future consequences; temporal distancing produces the perspective-taking effect that reduces present-state bias and engages the analytical processing of future outcomes). The 10/10/10 framework, developed by business writer Suzy Welch and grounded in the psychological research on temporal distance, asks three questions that force explicit consideration of the decision's consequences at three time scales — reducing the dominance of the immediate emotional response and activating the longer-term analytical processing that present-state bias suppresses.

Implementation steps:

• For any decision with significant long-term consequences, answer in writing: "How will I feel about this decision in 10 minutes?" (captures the immediate emotional response)
• "How will I feel about this decision in 10 months?" (forces consideration of medium-term consequences and the adaptation that present state makes hard to anticipate)
• "How will I feel about this decision in 10 years?" (surfaces the values and priorities that present emotional state consistently underweights)
• If the 10-minute answer and the 10-year answer conflict substantially, treat that as a signal that the affect heuristic may be driving the immediate response — and weight the 10-year perspective more heavily in the final decision

Quick Win — The 10-Minute Pre-Decision Checklist

For one significant decision you are currently facing, run through this five-question pre-decision checklist before committing:

1. Affect check: Do I have a strong emotional pull toward or against one option? If yes, have I specifically examined whether the emotional response is tracking actual evidence or the affect heuristic?
2. Base rate check: What typically happens in situations like this one? (Not what I expect to happen — what base rates suggest happens to people who make this decision in these circumstances.)
3. Pre-mortem check: If this decision fails, what are the three most likely reasons? Are any of those failure modes preventable?
4. Confidence calibration: On a scale of 1-10, how confident am I that this is the right decision? When have I been this confident before? Was I right?
5. Regret minimization: Which decision would I more regret at 10 years — choosing this option and having it not work, or not choosing it and wondering what would have happened?

See also: How to Stop Procrastinating for Pychyl's research on emotional avoidance as the root cause of delayed decisions, How to Set Goals and Achieve Them for Oettingen's WOOP framework applied to goal commitment, How to Develop a Growth Mindset for Dweck's research on how fixed vs. growth attribution affects decision-making after failure, and How to Increase Productivity for the full output architecture that creates the conditions for high-quality thinking.