Communications of the ACM
Lousy Advice to the Lovelorn
Credit: Akuma-Photo
People seeking advice on getting married have recently been directed to a decidedly unromantic algorithm. The algorithm has two stages. Let N be the maximum number of people you expect to be able to date before you give up. In the first stage, you date and dump N/e ≈ .37N people to get a sense of the overall quality of the field. In the second stage, you continue down the list, and you marry the first person that is better than everyone you met in the first stage. (If you reach the end of the list, and the last person is not the best, then the algorithm is indifferent; you can marry them or not.) There is a theorem that supposedly states that following this strategy maximizes the likelihood of marrying the best possible partner of those on your original list.
This algorithm is the subject of the first chapter of Algorithms to Live By, by Brian Christian and Tom Griffiths,1 and has been repeated in recent articles in The Washington Post,6 Business Insider,1 Slate,5 and NPR.4 The headline in Business Insider makes the advice even simpler: 26 is the perfect age to get married.
Comments
Roland Kwee
Thanks for your clear analysis of this fun topic. However, the false presumption is that one should look for the "best" partner. Indeed a modern disease. To increase the chance that your genes will survive, and to live longer, healthier, and happier, you should only look for the minimal qualifying partner. The economic theory of diminished returns suggests that the effort for finding another partner that is better than the previous one might outweight the advantages of that better partner. And the article described the bigger chance to end up with no partner at all, or be forced to propose to one that is definitely not the "best", if you are too selective. Perhaps your article could have been shorter, and there are many other, more useful, math riddles.
Displaying 1 comment