| Chapter 2 Sunk costs: costs that were previously invested and that are nonrefundable Honoring sunk costs: behaving as if the nonrefundable expense is equivalent to a current investment Violates the first criterion of rationality: decision should only be based on future consequences Best way to avoid sunk costs is to start from the present and look to the future Probabilities can be assessed properly only with reference to future events Gambler's fallacy: when events are independent you shouldn't take the past into account People expect chance sequences to be locally representative of chance, whereas the true probability comes out only over an infinite amount of trials It's different to assess the probability of an entire pattern and that of an independent event given past events Law of small numbers: incorrectly assume that a small sample is as representative of the general population as a large sample Chance is not self-correcting; it just dilutes high scores with scores closer to the average It is harder not to honor sunk costs if other are involved Reputation and self-concepts may play a role Children and non-human animals are less prone to honoring sunk costs We have a tendency to consciously overgeneralize rules and to be limited by our conscious capacity to handle information Automatic processes may be able to handle more information at one time To avoid traps: Set limit in advance whenever it's possible Use limits as a time to reassess whether withdrawal or persistence is wiser, independently of prior investment; it must be the result of future-oriented cost-benefit analysis |
 |