You make a decision. It turns out badly. You conclude you made a bad decision. This is one of the most common — and most costly — reasoning errors humans make. It’s called resulting, and poker players learned to root it out long before behavioral economists gave it a name.
The Core Insight From Annie Duke
Annie Duke’s 2018 book Thinking in Bets argues that most of us conflate the quality of a decision with the quality of its outcome. A bad outcome from a good decision is just bad luck. A good outcome from a bad decision is also just luck — and it’s more dangerous, because it reinforces the wrong process.
Related: cognitive biases guide
Professional poker is a forcing function for this distinction. In the long run, good decisions produce better results than bad ones. But over any short series of hands, luck dominates. The discipline of separating process from outcome — “did I make a good bet given what I knew?” rather than “did I win?” — is what separates winning players from losing ones over thousands of hands.
Resulting in Real Life
Consider a driver who runs a red light and makes it through safely. They conclude: not a big deal, I do this sometimes. This is resulting — using the outcome to evaluate the decision. The decision (run a red light) was poor regardless of outcome.
Or a student who crammed the night before an exam and happened to get a good score. They conclude: cramming works. This is also resulting. They got lucky on what appeared on the test. Their study process was still low-quality relative to distributed practice.
I’ve made this error teaching. Early in my career I tried an unstructured discussion format with a class that happened to go brilliantly. I repeated it with different classes and it bombed repeatedly. The first success was partly luck — that class was unusually engaged. I had to learn to evaluate the method, not the result. [1]
What Calibration Means
Duke emphasizes calibration — the alignment between your expressed confidence and your actual accuracy. A well-calibrated person who says they’re 80% sure about something is right about 80% of the time across many such claims. Most people are dramatically overconfident.
Philip Tetlock’s decades of research on expert forecasting (summarized in Superforecasting, 2015) found that calibration is learnable. The key practices: express beliefs in probabilities rather than certainties, keep score on your predictions, and update beliefs when evidence changes rather than when you feel embarrassed to have been wrong.
Three Tools to Think Better About Decisions
1. The 10-10-10 Frame
Before any significant decision: how will I feel about this in 10 minutes, 10 months, 10 years? This expands the time horizon and reduces the weight of immediate emotion on the decision.
2. Pre-Mortem Analysis
Before executing a decision, imagine it’s a year from now and things have gone badly. Work backward: what went wrong? This surfaces hidden risks without the ego defense mechanisms that activate after failure.
3. Decision Journaling
Write down significant decisions, your reasoning, and your confidence level at the time. Review periodically. This creates an honest record that can’t be revised by hindsight bias — and it’s the fastest way to identify your actual patterns of systematic error.
Last updated: 2026-05-19
About the Author
Published by Rational Growth. Our health, psychology, education, and investing content is reviewed against primary sources, clinical guidance where relevant, and real-world testing. See our editorial standards for sourcing and update practices.
Your Next Steps
- Today: Pick one idea from this article and try it before bed tonight.
- This week: Track your results for 5 days — even a simple notes app works.
- Next 30 days: Review what worked, drop what didn’t, and build your personal system.
Sources
References
- Duke, A. (2018). Thinking in Bets: Making Smarter Decisions When You Don’t Have All the Facts. Portfolio/Penguin. Link
- Duke, A. (2018). Thinking in Bets: Making Smarter Decisions When You Don’t Have All the Facts. Big Think. Link
- Duke, A. (2018). Thinking in Bets by Annie Duke | A Guide to Smarter Decisions. YouTube – Clay Finck. Link
- Duke, A. (2018). Beyond Luck—Behavioral Science and the Art of Decision Making. T. Rowe Price – The Angle Podcast. Link
- Duke, A. (2018). Book Summary: Thinking in Bets by Annie Duke. The Exceptional Skills. Link
- Duke, A. (2023). Thinking in Bets for Engineers — with Annie Duke. Refactoring.fm / YouTube. Link
Related Posts
- Skin in the Game: Why You Should Only Trust Advice From People
- Ultralearning by Scott Young [2026]
- Why Unfinished Tasks Haunt Your Brain [2026]
The Cost of Outcome Bias in High-Stakes Professions
Resulting isn’t just a personal reasoning flaw — it has measurable consequences in fields where decisions carry serious weight. A 2003 study by researchers Hal Arkes and Cindy Schipani examined how outcome knowledge influenced mock jurors evaluating medical malpractice cases. Jurors who were told a procedure led to a bad outcome rated the physician’s decision as significantly more negligent than jurors assessing the identical procedure when the outcome was neutral — even when the medical reasoning was identical in both cases. This is outcome bias in legal form, and it has real effects on how professionals are judged and how they subsequently behave.
In financial advising, the same distortion appears. A 2012 paper by Rüdiger Weber and Colin Camerer found that fund managers were disproportionately likely to be fired following periods of underperformance even when their process was statistically sound — and that replacement managers hired after poor runs rarely outperformed their predecessors over the following three years. The terminations were decisions driven by outcomes, not evidence of underlying skill decay.
Emergency medicine offers a useful counter-model. Trauma teams increasingly use structured debriefs that explicitly separate outcome from process: a patient who dies despite a correctly executed protocol is documented differently from one who dies following a procedural error. Massachusetts General Hospital’s simulation training program, cited in a 2019 review in Academic Emergency Medicine, found that teams trained to debrief using process-first language showed a 23% improvement in protocol adherence over 12 months compared to teams using outcome-first review. The discipline of asking “did we execute well?” before “did it work?” produces measurable learning gains.
Probabilistic Thinking and the Overconfidence Gap
Most people don’t think in probabilities naturally, and the gap between stated confidence and actual accuracy is larger than intuition suggests. In Tetlock’s Good Judgment Project — a forecasting tournament that ran from 2011 to 2015 and involved roughly 20,000 participants — the average forecaster showed a confidence-accuracy gap of about 15 percentage points on binary geopolitical questions. When participants said they were 75% sure of something, they were right closer to 60% of the time. The top 2% of forecasters, labeled “superforecasters,” closed that gap to under 3 percentage points by using three specific habits: expressing all beliefs in numerical probabilities, actively seeking disconfirming information, and updating predictions in small increments rather than wholesale revisions.
The same overconfidence pattern shows up in business planning. A study by Bent Flyvbjerg at Oxford, drawing on data from 2,062 infrastructure projects across 104 countries, found that cost overruns averaged 44.7% in real terms and schedule overruns were present in 86% of projects. The primary driver was not technical failure but optimism bias — project teams systematically assigned low probabilities to delays and cost inflation that historical base rates predicted with high reliability. Flyvbjerg’s remedy, which he calls “reference class forecasting,” is essentially applied calibration: before estimating project outcomes, look at the distribution of outcomes for the 50 most similar past projects and use that distribution as your prior.
For individuals, a simple calibration practice requires no special tools. Keep a decision journal: record the decision, your confidence level (expressed as a percentage), and the reasoning behind it. Revisit entries after 90 days. Most people discover within three months that they are systematically overconfident in a specific domain — career predictions, relationship assessments, financial projections — and underconfident in others. That asymmetry, once visible, is correctable.
Building a Personal Decision Audit System
One underused application of poker-style thinking is the retrospective decision audit — a structured review that evaluates past choices on process rather than outcome. Gary Klein, a cognitive psychologist who developed the pre-mortem technique, also described a complementary “decisional autopsy” process used in military planning contexts: after an event, analysts separately evaluate (1) what information was available at decision time, (2) what the decision-maker could reasonably have inferred from it, and (3) what actually happened. Only after steps one and two are documented does the team examine step three. This sequencing prevents the outcome from contaminating the process evaluation.
Adapted for personal or organizational use, a lightweight audit system involves four questions logged within 48 hours of any significant decision:
- What did I know at the time, and what did I not know?
- What was my stated confidence level, and why?
- What alternatives did I actively consider and reject?
- What would have needed to be true for a different choice to be correct?
Research on structured reflection supports the practice. A 2014 study by Giada Di Stefano and colleagues at Harvard Business School found that employees who spent 15 minutes at the end of a work day writing structured reflections on their decisions performed 23% better on subsequent problem-solving tasks than a control group that simply continued working. The mechanism appeared to be consolidation: writing forces articulation of reasoning that otherwise stays implicit and unexamined. The audit doesn’t require much time — it requires the discipline to do it before you know the outcome.
References
- Arkes, H.R., & Schipani, C.A. Medical malpractice v. the business judgment rule: differences in hindsight bias. Oregon Law Review, 2003.
- Flyvbjerg, B. What you should know about megaprojects and why: an overview. Project Management Journal, 2014. https://doi.org/10.1002/pmj.21409
- Di Stefano, G., Gino, F., Pisano, G., & Staats, B. Learning by thinking: how reflection aids performance. Harvard Business School Working Paper, 2014. https://www.hbs.edu/faculty/Pages/item.aspx?num=46893
