Dating: You’re Doing It Wrong

In ‘Data, A Love Story,’ Amy Webb details her almost unbelievably successful quantified dating experiment. She started online dating with terrible early results — one bad date after another, including one who walked out on a $1,300 dinner tab. The first thing she started tracking was occurrences of bad date behavior, like high fives and inappropriate sexual comments, just to prove to her mother that she wasn’t being picky. She soon had a pretty good list of attributes she didn’t want, but what did she want? She made a list. A very long list. She turned this list into a scoring system to empirically evaluate whether she should even go on a first date.

Webb found that she wasn’t getting responses from the men who scored high, so she used data to evaluate her own profile. She set up a few fake male accounts, and used them exclusively to scrape data. She discovered the formula that makes women popular on dating sites: show some skin, smile, use the word “fun,” and don’t write more than about 100 words in your profile. She then optimized her profile, and immediately received a flood of responses. But none of these men met the benchmark of 700 points on her scoring system, so she didn’t agree to any first dates. The first man to hit the mark started with 850 points and went up to 1,050 after a 14-hour first date. They are now happily married with a daughter, and she never had to go on another bad first date. I highly recommend her hilarious TED talk.

She’s not the only one to game the dating system using data. Mathematician Chris McKinlay used mathematical modeling and an army of intelligent bots to outdo OkCupid’s algorithms. The site wasn’t working for him, and he decided to see if he could get better results. He was able to lump women on the site into seven clusters, he then optimized two different profiles for the two groups he liked. He used text analysis to find common keywords and interests and wrote his profiles accordingly. He picked the 500 most popular OkCupid questions and answered all of them, then he created a program to visit the pages of his top matches, so they would get notifications. The messages started pouring in at over 20 per day — enough that he could be picky.

McKinlay wasn’t as successful as Amy Webb. It took 88 first dates for him to find a girlfriend, but after months of rejection and no dates, his experiment proved that data analytics truly can improve one’s dating odds. Productivity geeks use personal data for lifehacking purposes all the time, so why not hack dating?

The author of “Quantified Breakup” didn’t end up with the man of a thousand texts. Like many relationships that start with sparks, this one fizzled in a few weeks. But she isn’t upset about that. She has rediscovered her ability to connect, and with the data from her first few dates, she’s better equipped to focus her search. The fact that she sent a bazillion texts to a guy she likes may seem like a pretty obvious correlation, but it helps tell a story. It provides a concrete way to quantify a fleeting emotional response.

Gathering data on dating is a way to approach an ambiguous and emotionally fraught process more logically. I don’t need an algorithm or a record number of text messages to know if I like a guy, but a few thoughtful calculations — and a heap of self-knowledge — might help me decide whether to go on that first (or second) date. The fact is, a crush is not enough on which to build a relationship. Chris McKinlay and Amy Webb aren’t too picky, they’re smart about finding the right match. Treating dating as a game or experiment (at least in the early phases) might make the arduous process a little more enjoyable, and help us avoid bad dates and unhappy endings.