Repelling the availability bias calls for an empirical approach to a particular decision, one not based on the obscured reality of vivid memory. If there is a low incidence of disaster, like only one out of 100,000 plane landings results in a crash, it is safe to fly to the Canary Islands. If one out of one million people who go out is assaulted, it is safe to go out at night.

The sunk cost fallacy has a periodic application and was first identified by economists. A good example of how it works is the casino slot machine. Gamblers with a high threshold for risk put money into a slot machine hoping for a big return, but with each pull of the lever they lose some money playing the odds. If they have been pulling the lever many times in a row without success, they might decide that they had better keep spending money at the machine, or they will have wasted everything they already put in.

The truth is that every pull of the lever has the same winning probability of nearly one in a trillion, regardless of how much money has been put in before - the previous plays were sunk costs.

In everyday life this can lead people to stay in damaging situations because of how much they have already put in, stuck on the erroneous belief that the value of that time or energy they have invested will decay or disappear if they leave. The wisest course is to recognize the effects of the sunk cost fallacy and to leave a bad situation regardless of how much you have already invested.

While there are still more biases, the key to avoiding them remains the same: When a decision matters, it is best to rely on watertight logic and a careful examination of the evidence and to remain aware that what seems like good intuition is always subject to errors of judgment.