I've always wondered why psychologists spend time and money on studies that show things every man who has ever been to a bar knows. Also, is there evidence to think that the behavior described here is not irrational at all. If less attractive people choose partners based on features other than appearance, and they do so based on (1) an estimation of their own attractiveness, and (2) a recognition that their own attractiveness is likely to limit the range of potential partners, then their behavior is paradigmatically rational. Perhaps the "irrationality" here is meant to be under the surface in some sense (though there could be rational, but not explicitly conscious, processes); but I suspect that most people make these decisions consciously.

Here is what would make this data interesting: are the people in this study looking generally for short-term relationships, or long-term commitments? Do the selection criteria vary depending on which they are seeking? For example, if attractive people always privilege appearance as a selection criteria, but it turns out that relationships based on appearance are less likely to be successful, then the conclusion would be that attractiveness is detrimental to long-term rationality.

Why is this irrational? Why is this interesting?

Ariely doesn't quite raise the irrational aspect to the level of explicit discourse, but the fact that people seek a partner somewhat more attractive than themselves leads to unanticipated consequences. If everyone knew their level and were content to have someone at their level, then everyone could be happy. But if everyone knew their level and wanted someone a slight step up, nobody could be happy. Imagine two ladders. Each person wants to join hands with someone higher up on the other ladder, but since the other people are trying to do the same things, nobody joins hands. Add to that the fact that age diminishes attractiveness - people are slowly sliding down their ladders - and you have an even more irrational situation. Since you're going to be less attractive tomorrow than you are today, it might pay to pick someone a little less attractive if you think they'll age well, and get into a long term commitment.

When you look at some of the research from long ago, you actually find that more attractive people become pickier faster than they become more attractive, so that there is no net benefit to being more than average. People in the bottom group of attractiveness in one famous study (Berscheid and Walster) had a real disadvantage in the mate market, but those in the top group had become so picky that they didn't have any better chance of making a happy match than people in the middle group. This is a famous paper, and nobody but me, to my knowledge, has cranked through the matrix algebra to arrive at this conclusion. The study was done in the early 1960's and is one of the soundest methodologically (it was a randomized, blind dating study of college freshmen).

This work is important and interesting because so much social discourse tries to tell people that beauty is only skin deep and in the eye of the beholder. Although that is partly true, there are other phenomena less democratic and more influential in social life. Understanding what really is (instead of what we want to be) could revolutionize, let's say, the psychotherapy of relationships. This research might ultimately lead people to adopt a more reasoned and humane set of expectations. Relationships are almost always a compromise, and if both partners understood that the other will sometimes wish to stray - that this is human nature - we might all be happier and more forgiving.