Nair mentioned in the comments here something called ‘asymmetric dominance’. He mentioned this in the context of this talk by Dan Gilbert, the researcher who I pointed out had very interesting things to say about people and how we think about probabilities.
I want to try and explain what asymmetric dominance is. Let’s say that there are two versions of a watch (or car, or unicorn) that are on sale. The inexpensive version has fewer things of luxury, obviously. In this case, a survey of what people buy usually shows that the trend observed is correlated to the spending power. That is, if people are prepared to spend only so much on a watch, and only the inexpensive version fits in their budget, that’s what they’ll buy. If people can afford the expensive version, they buy that.
Let’s say now, that somebody (who simply couldn’t help themselves) introduces a third version of the watch. This version is priced somewhere between the other two versions. Except, on the scale put in place by the other two versions (i.e. difference in worth between the two versions, in comparison with the difference in price), this third version is hugely overpriced – it is only a little better than the inexpensive version, but is a lot more expensive. An explanation of the phenomenon of asymmetric dominance works best if this third version is patently and obviously bad, so let’s assume it is.
One would expect, rationally, that people will still only buy the two original versions. One would also expect, then, that the number (or ratio) of people who buy one version or another shouldn’t change. One would be wrong. It turns out that people, faced with this new third choice which is so obviously bad, choose the (original) expensive version a lot more often than they would have if they only had the two original choices.
‘Intelligent’ supermarkets, then (in addition to putting dairy products and bread and eggs as further back on their aisles as possible), put validly expensive products next to the third kind of obviously bad bargains, apparently.
A seemingly related observation is that people (including professional wine-tasters) who are asked to test a bottle of wine think it tastes better if they are told that it costs more than it actually does. Go figure.