Just a quick follow-up on the Dan Ariely lecture last night. He had a very good way of explaining information cascades (and how apparently 'popular' choices could be based on very little information). He gave the example of when you are trying to choose a restaurant to eat in on holiday. If there are two places to eat in a popular tourist spot, but neither has people in it, the first person to choose will make an uninformed choice. However the next tourist to choose will see that one restaurant has one person in it, and the other has no-one in it. So he will likely choose the former. The third tourist will see one restaurant with two people in it and one with no people in it, and again is likely to choose the more popular option. After this has taken place several times one restaurant will look very popular and the other will not. Yet this very clear trend is not based on relevant information (ie quality of the food), but rather is simply the result of the uninformed decisions of the first person to choose.