Remembering past choices as better than they were, while remembering rejected alternatives as worse.
Imagine you picked the red lollipop instead of the blue one. Later, when you think back, you remember the red one tasting extra amazing and the blue one looking kind of gross—even though they were almost the same. Your brain is like a friend who always tells you 'Great choice!' and rewrites the story so you never feel bad about what you picked.
Choice-supportive bias is a memory-level distortion that occurs after a decision has been made, in which people systematically misremember the attributes of chosen and rejected options to favor the choice they made. Positive features tend to migrate in memory toward the chosen option—even features that originally belonged to the rejected alternative—while negative features drift toward the rejected option. This bias operates largely through faulty source monitoring at the time of retrieval rather than through biased encoding at the time of the decision itself. The effect is robust enough that even giving people false reminders about which option they chose produces memory distortion in favor of the believed-chosen option, demonstrating that it is the belief about one's choice, not the actual choice process, that drives the distortion.
The same glitch looks different depending on the terrain. Finance, medicine, a relationship, a team — same mechanism, different costume.
Investors tend to remember their purchased stocks as having been clearly superior picks at the time, while retroactively attributing more flaws to stocks they passed on. This distorted memory impairs the ability to conduct honest portfolio reviews and learn from losing trades.
Physicians who chose a treatment protocol tend to recall more benefits of the selected approach and more risks of the alternatives when reflecting on outcomes, which can bias case reviews and impede evidence-based updating of clinical practice.
Mara Mather, Eldar Shafir, and Marcia K. Johnson formalized choice-supportive memory distortion in 2000, building on earlier cognitive dissonance work by Jack Brehm (1956) on post-decision attitude changes.
Constantly second-guessing past decisions would have been cognitively expensive and paralyzing for ancestral humans who needed to act decisively in survival contexts. A mechanism that retroactively reinforced confidence in past choices would reduce rumination, preserve mental energy for present threats, and maintain the social confidence needed to lead or persuade others in group settings.
LLM agents exhibit choice-supportive bias when acting as evaluators: after making an initial selection, they systematically attribute more positive features to the chosen option and more negative features to alternatives, even when no factual hallucination is present. A 2025 study across 19 LLM models found this bias increases when agents perceive they are in control of the decision, and it persists across diverse domains and prompt constructions.
This page taught you the name. The deck turns the name into reflex. 1,100+ swipeable scenarios, 1,100+ defenses, 650+ detection prompts — spaced-repetition Swipe Deck, unlimited Spot-the-Bias Quiz, Defense Playbook, Pre-Flight, My Blindspots, Cheat Sheets, Field Guide e-book. $39.53$59.
