A Good Decision Can Still Have a Bad Outcome

A Good Decision Can Still Have a Bad Outcome

A good decision uses the information available at the time, considers credible alternatives, accounts for uncertainty, and respects the right constraints. It can still produce a bad outcome. A bad decision can get lucky.

That does not mean outcomes are irrelevant. It means one result is evidence, not a complete verdict on the process that produced it.

This is the idea from Annie Duke's Thinking in Bets that stayed with me after the episode.

The book makes uncertainty easier to admit

Annie Duke spent years as a professional poker player before writing and consulting about decision-making. Poker gives her a strong setting for the argument because every player has incomplete information and because luck remains present even when a person plays a hand well.

Most decisions outside poker are not wagers in a literal sense. They still commit time, money, attention, reputation, or opportunity to one possible future instead of another.

Thinking in bets is a way to say, "This is what I believe, this is how confident I am, and this is what could make me change my mind." That language is more honest than pretending an uncertain choice has become certain because someone needs to act.

Resulting confuses the decision with the result

Duke calls it resulting when someone uses an outcome as the main proof that the preceding decision was good or bad.

The cleanest example is driving after drinking and arriving home safely. The lack of a crash is a favorable outcome. It does not make the decision safe.

The reverse mistake also happens. A person can follow a careful process, choose the strongest option, and still land in the part of the probability distribution where things go wrong.

Separating process from outcome protects learning in both directions. Success does not erase bad reasoning, and failure does not automatically reveal bad reasoning.

The Super Bowl example needs the correct game state

The publisher introduces the book with Pete Carroll's decision near the end of Super Bowl XLIX. Seattle trailed New England 28 to 24, had second-and-goal at the one-yard line, and had 26 seconds remaining. Russell Wilson's pass was intercepted by Malcolm Butler.

The episode mistakenly says Seattle was ahead by four with several minutes left. That changes the decision context and should not be repeated.

Duke's purpose is not to prove that every controversial decision was correct. It is to ask whether people are evaluating the information and options available before the play or merely reasoning backward from the interception.

That distinction is useful far beyond sports. A review should reconstruct the decision state before it lets the known outcome dominate the room.

A process can be reviewed without pretending it was perfect

Calling a decision process sound is not immunity from accountability. The process may have ignored evidence, omitted a stakeholder, underestimated a foreseeable harm, exceeded someone's authority, or used a model that was never validated.

Outcome evidence can reveal those weaknesses. Repeated outcomes can show that a probability estimate was poorly calibrated or that a risk thought to be rare was not rare in practice.

The fair standard is to ask two sets of questions.

First, was the decision reasonable given what was known and knowable at the time? Second, what does the result teach us now that we have more evidence?

Truth-seeking needs a conversation contract

The episode's most human section is about groups that challenge one another's reasoning.

A useful review group does more than confirm that someone was unlucky. It asks what the person believed, what alternatives were considered, which information was missing, and what would change the decision next time.

That style of conversation is not appropriate every time someone shares a problem. A friend may want to vent. A colleague may need support before analysis. Someone may not have invited a full review of their choices.

Dalton calls this the conversation contract. Before offering hard analysis, find out whether the person wants comfort, a sounding board, advice, or a decision review. Truth-seeking without consent can become another way of making the conversation about yourself.

Long-term thinking is easier when the choice is designed

The episode uses the Seinfeld idea of Night Jerry and Day Jerry to describe a conflict between the present self and the person who will live with the consequences tomorrow.

The practical response is not endless self-control. Change the choice architecture. Put the phone away from the bed. Automate a contribution. Arrange transportation before drinking. Create the constraint while the long-term goal is still clear.

That idea complements thinking in bets. A decision process should consider more than probabilities. It should also account for how attention, emotion, timing, and environment affect the person making the choice.

My review of the book

I liked Thinking in Bets, and I would recommend it to someone who wants a readable introduction to decisions under uncertainty.

The stories did most of the teaching for me. The poker hands, the executive decision, and the football example made the distinction between process and outcome easier to remember than an abstract rule would have.

My main criticism is that I wanted more of those stories. Some concepts also deserve careful sourcing before they are carried into a workplace method. A memorable metaphor is an invitation to examine a decision, not a validated model for every domain.

The book is strongest when it makes three phrases easier to say: I am not certain. This is what I believed at the time. Here is what the result changed.

Continue the conversation

The episode works through resulting, probability, truth-seeking groups, conversation boundaries, and the conflict between present and future choices. Listen on Spotify or watch on YouTube.

Sources and editorial notes

This article uses the preserved [[E15 - Transcript - ep15-bet-on-yourself-a-review-of-annie-dukes-thinking-in-bets|raw transcript]], the publisher record for the book, World Series of Poker records, and the NFL account of the Super Bowl XLIX interception. It is a book review and decision-process explainer, not financial, gambling, legal, medical, employment, insurance, or other professional advice.