Say there's a two-sided matching game, where you trade a good with another; not trading generates no utility, so you want to get the highest possible value. All goods, including yours, are of uncertain value, and you have a noisy, unbiased estimate of all the values, as do others. Some have less noise around their estimations than others, though you do not know who. So, when someone shows up, and really really wants to trade their good for yours good, are you excited? No. They probably they have value well beneath yours. What is most interesting is if the person is marginally interested in trading with you, because that means they might be right in the sweet spot: not so low you could do obviously better, but high enough to be practically optimal given search costs.
This is exactly what they find in dating markets, surprising no one. People are attracted more to those who are kind of interested, as opposed to really interested. This is consistent with the simple, rational bayesian model that is perfunctory for economists. It is almost more common than not that a popular book on economics caricatures this model as insanely unrealistic. We should remember that this model actually explains a lot, and when it fails often points out we are missing something in people's information sets, as opposed to something in their neurology. When a company announces positive news, the stock generally rises, when Treasuries fall, its usually because people are forecasting lower inflation or lower future real rates. These are uninteresting 'dog bites man' stories because they are so common and make sense using the standard rational model.