“So we don’t really know when the algorithm has worked.
On Tinder, a swipe to the left means you’re not interested and a swipe to the right means you like the person; if they “right-swipe” you back, it’s a match. He also wanted to test out his robotic invention — a mechanical finger that held a conductive pen. “Although,” he adds, “I’m not sure how many of those were robotic fingers.” He received about 70 messages from girls within a 100-mile radius, but he only replied to one girl who worked in engineering.He called it the “Tinder-O-Matic,” which “likes” a new profile every 4 seconds, or 900 likes an hour. “I thought she might get a kick out of the project,” he says.He was 100% upfront about his “Tinder-O-Matic” as an experiment and he says it’s less invasive than software hacks like “Tinder Auto Liker,” which claims to do the same thing.In the online dating arena there’s a reason people are increasingly turning to apps that rely on photographs and data instead of long essays about being a lover of Broadway shows and having a good sense of humor.“Dating algorithms do exactly what they’re designed to do, which is to match two sets for data,” says Amy Webb, author of “Data, a Love Story: How I Gamed Dating to Meet My Match.
“That doesn’t mean that the algorithms fulfill our intentions, which is to find a mate and to settle down into a long-lasting relationship.” Why do dating sites fall short?The questions posed by the various sites are too rudimentary, Webb says, and they tend to focus on a grocery list of requirements.“In order to really match people based on more than preferences — which is to say their personalities, their observable behaviors and the like — you’d need more sophisticated machine learning, more natural language recognition, more computing power,” she says.Location-based apps like Tinder focus on looks alone.Making sense of dating algorithms is a virgin science and is still more miss than hit, says Mark Brooks, a dating-industry analyst and the editor of Online Personals Watch.When people do find somebody they like, sites don’t often know why or how they’ve been successful, he says.