brain/
conceptpsychology

Resulting: judging decisions by outcomes rather than process

Notes

Resulting: judging decisions by outcomes rather than process

One-line summary: Outcome quality is a noisy signal of decision quality. Evaluating a choice by how it turned out (and taking credit for good luck while blaming bad luck) contaminates learning; Duke's fix is to deconstruct the process before the result is known, and to hide the result from reviewers who did not see it.

What it is

Thinking in Bets names this "resulting." This 2018 interview never uses that word; it states the same rule. Two mechanisms:

  1. Outcome contamination of process review. "If an outcome has already happened... you're going to try to reason to make the outcome make sense." Once the option has expired or the hand is over, the result rewrites the story of whether the decision was good. Duke's operational counter: deconstruct the trade before the result; if you are already past the outcome, do not tell reviewers who don't already know it.
  2. Self-serving bias as the default scoring rule. Nature's definition of "winning" is "affirming that our beliefs are already true, that bad things that happen aren't our fault, that good things that happen are to our credit." The bet-frame relocates winning to "having the most accurate representation of the world."

Poker makes both vivid because every hand has a marked P&L. Investing has the same structure with slower feedback.

Why it matters to psychology

This is the decision-science version of the outcome-bias literature (Baron & Hershey and after): luck is not a valid process grade. For a thread that cares about personal decision-making and Tetlock-style forecasting, resulting is the error that makes calibration impossible — you cannot update beliefs honestly if every good result confirms you and every bad result is "the environment."

Single-source, interview restatement of a book argument. No primary experimental citations in this episode.

Evidence

Implications

  • Decision reviews (investment memos, post-mortems, personal journals) should be timestamped before outcomes, or outcome-blind for anyone who wasn't in the room.
  • Self-serving bias is not a separate "being a jerk" problem; it is the scoring rule that resulting uses.
  • Group hygiene (group-decision-hygiene disinterestedness) is how you instantiate this when you cannot review yourself.

Open questions

  • How much outcome-blinding survives once market prices / public scoreboards make the result common knowledge?
  • Is process-only grading teachable, or does it require an external referee (the "bet" or a third person)?

Related

Referenced by