What Would Change My Mind?
MinutesBefore defending a belief, write down the specific evidence that would make you abandon it, which turns the belief into a testable claim and exposes positions (yours or a persuader's) that nothing could ever dislodge.
How to do it
- 1State the belief in one sentence and rate your confidence from 50 to 100 percent.
- 2Write the concrete observation, study, document, or event that would move you at least halfway toward the opposite. Vague answers ("real proof") do not count; name the source and what it would have to show.
- 3If you cannot write anything, notice that: a belief that no evidence could touch is not a belief about the world, it is a commitment, and it should be labeled as one.
- 4Ask the same question of the person persuading you. An honest advocate can answer in a sentence; a manipulator changes the subject, raises the bar every time it is met, or treats the question itself as an attack.
- 5Keep the answer where you will find it again. When the named evidence arrives, the cool version of you has already agreed what it means.
What to say
- “What would I have to see to conclude I was wrong about this? Let me write it down before I argue.”
- “What would change your mind? If the answer is nothing, then we are not discussing evidence and I will stop treating this as a debate.”
When to use it
- •Before arguing for a position you feel strongly about, especially one tied to your group or identity.
- •When a claim keeps surviving every counter-example by shifting what would count.
- •When evaluating a leader, group, or theory that explains every outcome, including the failures, as confirmation.
- •When you notice you are defending a conclusion rather than examining it.
Counters
Evidence and how strong it is
The question operationalizes Popper's criterion that a claim which forbids nothing explains nothing (Popper 1959), and it targets a well-documented failure: beliefs persist after the evidence that produced them is discredited (Ross, Lepper & Hubbard 1975), and people rarely specify in advance what would falsify them. Tetlock's forecasting research found that the forecasters who beat experts and prediction markets were distinguished by treating beliefs as hypotheses and updating in small steps when specified evidence arrived (Tetlock & Gardner 2015); Galef (2021) collects the practice as the central habit of what she calls a scout mindset. Evidence strength: strong for the mechanism (unfalsifiable beliefs resist correction; pre-specification improves updating), but the question as a stand-alone everyday tool is practitioner practice, not a tested intervention.
- Pre-specifying evidence is not a promise to change your mind the moment something resembles it. Weigh the evidence when it arrives; pre-commitment sets the terms, not the verdict.
- Some beliefs are values rather than predictions ("cruelty is wrong"); the question does not apply to them, and asking it can make a person feel their ethics are on trial.
- Used aloud in a hostile argument, the question becomes a rhetorical trap. Ask it of yourself first, and of others only when you would accept an honest answer.
- Popper, K. R. (1959). The Logic of Scientific Discovery. HutchinsonFalsifiability as the mark of an empirical claim, the principle the question puts into practice.
- Ross, L., Lepper, M. R., & Hubbard, M. (1975). Perseverance in Self-Perception and Social Perception: Biased Attributional Processes in the Debriefing Paradigm. Journal of Personality and Social Psychology, 32(5), 880-892Experimental demonstration that beliefs survive the discrediting of the evidence that created them.
- Tetlock, P. E., & Gardner, D. (2015). Superforecasting: The Art and Science of Prediction. CrownFindings from the Good Judgment Project that treating beliefs as revisable hypotheses and updating on specified evidence distinguishes accurate forecasters.
- Galef, J. (2021). The Scout Mindset: Why Some People See Things Clearly and Others Don't. PortfolioA practitioner synthesis that puts the what-would-change-my-mind question at the center of accurate reasoning.