Preference Weighting and the Abilene Paradox — LessWrong