I keep a notebook handy so that when I hear or read something interesting I can jot it down. And then later, when I’m supposed to be doing something else unpleasant, I’ll leaf idly through the notebook looking for a distracting tidbit.
Yesterday, for instance, I was avoiding paying my American Express bill and spotted this in my notebook:
“Studies show that relationships between people who met online are no more dependable and lasting than relationships that started between people who met at Starbucks or on the subway.”
That surprised me because I have always assumed that computer matching algorithms, the kind that Netflix uses to suggest other shows to watch or Amazon uses to recommend additional purchases, were pretty dependable predictors of human behavior.
Dating and matchmaking sites gather an enormous amount of information from their users — initial preferences, chat histories, number of matches, that sort of thing — and they manipulate and spindle all of that data to present their users with the most scientifically precise set of love matches available.
But for some reason, a serendipitous meetup in a local coffee shop has about the same odds of success. Random chance, in other words, is about as lucky as hundreds of computer scientists engineering thousands of lines of code. And that’s why, next to this little factoid scrawled in my notebook, I punctuated the entry with an enormous “?” symbol.
I also, apparently, looked up the average annual salary for a second-tier computer scientist at Tinder, the most popular dating app. It’s about $250,000 when you add in bonuses and stock options. Which seems a little rich for devising something about as accurate as saying, “Excuse me, can I just grab a lid?” at the local Peet’s.
My afternoon was derailed trying to figure this out.
Meeting someone in a haphazard way, in real life, takes a lot of social courage. Most people dread making the first move so that when someone actually does there’s a powerful romantic charge in the air.
And maybe a legal charge, too, these days, when saying, “Is this my latte or yours?” to an attractive stranger is considered a form of sexual battery. I think that’s why these kinds of connections are so powerful: Someone risked jail to get your number! That’s gotta mean something, right?
Romantic movies are all about what they call the “meet cute,” and it will take thousands of gender studies professors thousands of years to eliminate the little frisson human beings feel from a chance romantic encounter. The randomness confers the event with something deeper — it had to be karma, fate, the hand of God that led us to that same Chick-fil-A.
Maybe that’s why the relationships that come from these kinds of meetings are so lasting. The person may end up being annoying or unreliable or deeply weird, but Cupid can’t be wrong, so stick with it.
It’s the same with dating apps. People who meet online or via an app assume that the computer algorithm must know something they don’t. All of those computer scientists with those rich pay packages can’t be wrong, can they?
When two people swipe right on Tinder, the match has the authority of advanced mathematics going all the way back to Isaac Newton and Carl Friedrich Gauss. Those were pretty smart guys, and if they say you two should meet for a drink after work and see if there’s a vibe, you’d best do it. And when you do, you will try a little harder to see what the computer saw. To see, in other words, the best in the other person.
That’s why dating apps and meet cutes have roughly the same rate of success. In both cases, there’s a powerful unseen authority encouraging you to make it work.
Unlike, say, your mother or your co-worker, who have no idea what makes you happy. Their setups are embarrassing failures.
Full disclosure: We’re talking about a clipping that’s been in my notebook for at least a year. It may have been thoroughly debunked for all I know.
But it does suggest that there may be an opportunity to create a successful dating app that has zero technology behind it. It could just be a randomly generated pairing program, based on location, that combines the tingly sensation of the random meetup with the trappings of advanced data science. We’re talking maybe four lines of code. You wouldn’t need even one $250,000 computer scientist. All you need are people who believe that it’s up to them to make it work.
Rob Long is a television writer and producer and the co-founder of Ricochet.com.
