Lexie Perez
Graphic depicting two people finding their network of love on computers with a large heart with a question mark in it above them. (Hustler Multimedia/Lexie Perez)
“We were algorithmically matched with each other in college.”
That’s a new one! Think about it: Have you met a couple who can tell you they left their matchmaking to science and coding? Thanks to a continually evolving algorithm created by a group of Harvard students in 1994, Datamatch has provided an exclusive platform for college students to meet both platonic and non-platonic “friends.” Now, 30 years later, the online algorithm is rapidly gaining popularity, reaching over 40 colleges and matching thousands of students since its debut.
How does it work? The process is quite simple: Each year on Feb. 7 at 12:01 a.m. EST, a campus-specific survey is released based on the university email address you signed up with. Between then and Feb. 14, you can fill out a short (10-15 questions) survey. After that, it’s a waiting game until you are matched with about 10 other students on Valentine’s Day.
By default, Datamatch’s algorithm matches students based on the similarity of their responses to the campus-specific survey questions. However, users may opt to be matched with opposites by choosing the appropriate option on their profile.
Howard Huang, part of the Datamatch team at Harvard, provided us with a glimpse into the inner workings of the unique matchmaking site. As a member of the Statistics team, Huang highlighted the collaborative efforts of various teams — Web, Algorithm, Business and Design — that all play a pivotal role in shaping Datamatch’s success. From the initial paper surveys of 1994 to today’s more developed digital platform, Datamatch has continuously evolved, maximizing its matching capabilities.
Interestingly, Huang noted that the core of the matchmaking algorithm has remained relatively unchanged over the years, still relying heavily on survey responses. Certain features have been added to the site to account for new metrics, but Huang emphasized that the main part of the matching algorithm falls on ensuring everyone gets a reasonable amount of matches.
“Some of the newer things like MBTI, love language [and] Rice Purity Score, were never collected in the past,” Huang said. “But the main part of the matching algorithm that’s very important is that we ensure everyone gets a good amount of matches within their constraints.”
What’s unique about Datamatch is the campus-specific questions that make up the matching survey. A group of campus ambassadors come together each year to submit questions for their school’s survey. For yet another consecutive year, Vanderbilt’s satirical newspaper, The Slant, has taken charge of this initiative. Questions from this year’s survey ranged from “What game do you play during lectures?” to “What’s your favorite type of gambling?”
Guided by insights from Sam Sliman, a senior and Managing Editor of The Slant, questions were designed with a balanced distribution between humor, local flavor and romantic insights.
“We always want to make sure we have a certain number of romantic questions and a certain number of Vanderbilt questions,” Sliman said.
Success for Datamatch at Vanderbilt has historically been gauged by qualitative metrics like the number of student sign-ups — a direct indicator of the website’s reach within the campus community. However, this year’s engagement took a noticeable turn, yielding only 3,101 matches compared to last year’s 5,821 total matches.
!function(){“use strict”;window.addEventListener(“message”,(function(a){if(void 0!==a.data[“datawrapper-height”]){var e=document.querySelectorAll(“iframe”);for(var t in a.data[“datawrapper-height”])for(var r=0;r