Dating by Numbers

A popular NJ radio station is WKXW-FM (101.5), with an all-talk format on weekdays. (On weekends it plays oldies music.) I’m often tuned to it while driving. On weeknights Michelle Jerson hosts a call-in show called After Hours, on which she asks such relationship-oriented questions as (these are taken from her page at the station web site) “How do you break up with someone?” or “Did you (or would you) take your husband’s last name?” or “Do modern men like aggressive women?” and so on. What always surprises me is how persistent the traditional gender divide remains in the answers even as the socioeconomic realities shift – it’s enough to make an economic determinist despair. To be sure, there are always unconventional callers, but this would have been as true in 1956 as today.

As an example, while I was driving home a couple weeks ago, Michelle asked listeners if they would consider dating someone who was unemployed. The unemployed are, of course, a sizable population in today’s economy, and the downturn has hit male workers especially hard. She predicted a gender split in the answers and she got it. Only a handful of women callers were willing to consider an unemployed man, and even those few usually qualified the answer with something like “if it’s clearly a temporary condition.” Most simply ruled it out. Men callers, on the other hand, by and large couldn’t have cared less if a prospective date was employed, or, if so, at what. They had their minds on other assets.

This is exactly the same division I would have expected half a century ago, despite the radical changes in wages and employment since then. Male wages, for example, peaked in real terms in 1973 and since have collapsed 32%; in the same time, female real wages rose 44%. The result, though, was not to reverse traditional patterns of attraction (most men are still drawn by looks and they still attempt to draw with status), but simply to make larger numbers of men undateable. In her Atlantic Montlhy article “All the Single Ladies,” Kate Bolick remarks, “Recent years have seen an explosion of male joblessness and a steep decline in men’s life prospects that have disrupted the ‘romantic market’ in ways that narrow a marriage-minded woman’s options: increasingly, her choice is between deadbeats (whose numbers are rising) and playboys (whose power is growing).” Perhaps partly in consequence (though plenty of other factors are at work), more of us of either sex just remain single.

The persistence of the old ways is also demonstrated by the ongoing success of Ellen Fein and Sherrie Schneider who made a splash (and a fortune) back in 1995 with All the Rules, a book of dating advice for women. The advice is unabashedly traditional. “Make him work for it” is the gist. The authors dismiss widespread criticism that their rules are relentlessly un-PC; they work, they say, and that is that. All the Rules still sells well 16 years later. I have no idea whether the Rules really work or not. Nor do I know if complementary advice aimed primarily at men works – e.g. the column by Jeremy Nicholson, the self-styled “Attraction Doctor” who blogs at Psychology Today. I always approached the whole business in a far more haphazard fashion than that, which may explain why I’m single.

Fein and Schneider do not advise on whom to apply the Rules. Nor does Nicholson, so far as I’ve noticed. Given Ms. Bolick’s complaint, this may be an important question. Actually, there is an optimal strategy for making this choice that was published as long ago as 1966. It was first brought to my attention in a blog by evolutionary psychologist Satoshi Kanazawa, but since he is in the publishing doghouse for having posted an amazingly ill-considered blog, I’ll bypass him as a source and go directly to the original article by mathematicians John Gilbert and Frederick Mosteller of Harvard: “Recognizing the Maximum of a Sequence,” published in the Journal of the American Statistical Association (http://isds.bus.lsu.edu/chun/teach/7025/7025-reading/Gilbert.pdf ).

I don’t pretend to follow all of their analysis of the so-called “beauty pageant problem,” which applies not just to pageants, but to choices among dates, jobs, movies, or just about any other multiple offering. However, the strategy that emerges from it is simple enough. Start with a fixed number of candidates. (Yes, you need a fixed number, but it can be anything that suits you – ultimately it may be the maximum number of people you are willing to date before giving up altogether.) Next, 1) Reject the first 37% out of hand, and then 2) pick the very next date whose qualities are better than anyone in the first batch you’ve already rejected. That’s it. You’ll have a 37% chance of getting the best candidate. 37% sounds like lousy odds, but no other method produces a better result. The strategy assumes, probably correctly, that you can’t return to someone you passed over – that he or she will have moved on.

The strategy is mathematically sound. Nevertheless, it seems to me that if you’re coolheaded enough to play it, maybe romance isn’t your game anyway. I suggest counting cards at Blackjack instead.

The Beauty Pageant Problem