Follow us on Instagram
Try our daily mini crossword
Play our latest news quiz
Download our new app on iOS/Android!

When The Marriage Pact came to Princeton (and matched a pair of twins)

Marriage Pact.png
Sydney Peng / The Daily Princetonian

“Fill out this survey to get matched with your other half,” the Princeton Marriage Pact promised. For the 2,273 Princeton students who completed the survey, it was an intriguing enough promise from a matchmaking service new to the University this year — especially within the context of returning to campus during the COVID-19 pandemic.

Despite the eye-catching advertisements and the buzz surrounding the service, most students may know little about the background of the Marriage Pact, an exciting trend and source of gossip that came and went.


The Marriage Pact began as the brainchild of two students at Stanford University as an economics project and has since grown into a full-fledged company under the command of CEO Liam McGregor.

“We’d heard of what a marriage pact is, and we thought ‘what are the odds that the person you know ... is actually the best person you should make that agreement with?’ ... At a place like Princeton with 5,000 kids, what are the chances that you happen to know the best person for you? Really low! So the question was, can we find that person?” McGregor said.

The Marriage Pact may have begun at Stanford University, but it has since made it to colleges all across the United States — and now to Princeton University. Initially, Princeton’s Class of 2021 class government looked at the algorithm as a potential way to match seniors from Princeton with seniors from Yale for friendship, but when the announcement came that Princeton’s undergraduates would be returning to campus, they decided to focus their attention solely on Princeton’s student body. Since every college is different, each iteration of the Marriage Pact is different. But its primary value remains in its community building.

“We ... hoped that it would be something that the campus buzzes about, or like a collective shared experience that people can gossip about together, like just have something in common, like a shared experience,” said Class of 2021 social chair Phoebe Park.

In the process of bringing the Marriage Pact to Princeton, the 2021 class government involved the 2022 and 2024 class governments, going through most of the planning and organizing process only a week before they rolled the program out. The 2023 class government chose not to be involved with the Marriage Pact.

The Undergraduate Student Government’s (USG) worked with the Marriage Pact to market it specifically for the Princeton community — using tailored advertisements that were sent through as many platforms as possible.


The Marriage Pact algorithm uses a participant’s answers to their survey to match them up with what the algorithm deems as their “perfect match.” The questions, a combination of tried and true questions already built into the algorithm and a few more Princeton-specific questions, cover everything from politics to TigerConfessions.

Every iteration of the Marriage Pact also comes with its own problems. At Princeton, most of those problems arose from the survey questions — primarily two that were built into the algorithm and that USG was not privy to before the survey’s launch.

One question asked if participants would support their child if they came out of the closet, and the other asked if their partner gaining weight would be a deal-breaker. Concerns were raised by some students about these questions for being insensitive, and the questions were swiftly removed from the survey.

“[Marriage Pact] doesn’t shy away from some of the heavier topics because they’re important to know about if you’re going to make a credible long-term relationship. The presence of a question on the questionnaire is not an endorsement of any particular position on it. I think it’s simply an acknowledgment that people do hold different views across these spectrums, and they’re important to know about for long-term relationships,” McGregor said.

Get the best of ‘the Prince’ delivered straight to your inbox. Subscribe now »

“I think there are ways we could’ve articulated questions differently. However, I do think that there are certain biases that I, as a gay Asian man, would like to avoid, and [the original Marriage Pact survey] did its job in helping me avoid that. And those questions were — as much as I understand why people found those questions to be problematic — I also appreciate the fact that I wouldn’t get someone with starkly different values based on the algorithm,” said Elliott Hyon, the 2024 class council Marriage Pact representative.

Aside from those concerns, the general reception toward the Marriage Pact was positive. It drummed up excitement and provided students with gossip and a fun anecdote to tell.

“I thought it was something interesting to do,” Rowen Gesue ’24 said. “It was cool. ... It did give a sense of community, a little bit, for something that other people were doing as well, which is nice during weird times like this.”

Fast forward to 7:49 p.m on Feb. 3, 2021, match day, just a few hours before the actual matches were released, a tweet about Marriage Pact gained some traction.

“got matched with my twin brother in the princeton marriage pact. how’s your wednesday going?” tweeted Andrew White  ’23, who entered the Princeton Marriage Pact with relatively small expectations and left with an unfortunate tale to tell.

When the Princeton Marriage Pact released the initials of participants’ matches, White and his twin brother received each other’s. Assumptions brewing, about an hour later, they found out that they had indeed been matched by the algorithm.

What’s more? The match was in the 100th percentile.

“I was like wow, a hundred percent, that’s something, isn’t it?” White said. “I mean, we are identical twins, we do similar things, we have similar personalities, so I guess it kind of makes sense that ... we would be matched by an algorithm. It was unfortunate, to say the least.”

At the same time the White twins were matched, over 2,000 other Princeton students were also opening their emails, anticipating the names of their matches. A few minutes prior to the announcement, a group of students unaffiliated with the Marriage Pact or class councils sent out a prank email, “Rick-rolling” the listservs by pretending to be Marriage Pact.

A few minutes after the real announcement, Tigerbook (a platform where students can look each other up) crashed, likely from the sheer amount of users searching for their matches. Some matched with people they knew. Some with people they didn’t. The Marriage Pact’s blunder in matching White with his twin wasn’t the only one made that night. Some received the wrong initials at first. Some matched with people whose values were too fundamentally different from their own.

Through aggressive marketing, the Princeton Marriage Pact managed to reduce the deficit of women to men in their heterosexual dating market from 161 unmatched women to closer to 30, but some of the women who signed up on the last open day of the Marriage Pact still ended up matched with a female friend rather than a potential romantic partner.

Not many people took the Marriage Pact survey expecting to find a future spouse, and most agreed that the Marriage Pact was a fun, unique experience, but its user base does have a few criticisms to lobby. For one, White believes that the problem he faced could have been solved if the Marriage Pact had looked more carefully at edge cases in their algorithm.

“I think that [the possibility of getting matched with your sibling is] an edge case that maybe should’ve been thought through,” White said. “I’m assuming that you can run a fairly quick check to make sure people don’t have the same last name. Otherwise, you could put in a ‘do not match’ list, like maybe just one person you’re afforded, so you can’t be too picky or selective, but like, one person where you can be like ‘I don’t want to be matched with this person.’”

Another participant with criticisms about the algorithm and survey questions was Orli Epstein ’24, though she enjoyed the experience overall.

“I thought the questions were not what you would ask if you were actually looking for someone to date,” Epstein said. “There were just a lot of irrelevant questions, and they didn’t ask anything that I thought mattered, for the most part.”

Ahmed Farah ’22 agreed that the survey could have been stronger, noting that when an algorithm strives to pair people based on their values, it’s likely that they will end up with a close friend or someone so similar to themselves that they won’t necessarily be compatible for a romantic relationship or, as in White’s case, a sibling. However, Farah also sees the value in Marriage Pact’s algorithm.

“It’s kind of refreshing to see something like this where you get to meet someone that you potentially could’ve spent four years at Princeton and never crossed paths with,” Farah said. “And the fact that that’s done through some method that tries to optimize for compatibility in terms of whether or not you two are going to get along or feel like you can actually hang out and understand one another because you have some sort of shared ground is, I think, really cool, and there’s definitely a space for it in college. I’m not sure if dating is the right space for it, but I think it’s a really cool thought.”

Members of class council also note that the Marriage Pact is more about meeting new people and bringing the community together than specifically about finding someone to spend the rest of your life with. And at the very least, they hope that participants in the Marriage Pact took something positive away from the experience.

Related: Can this algorithm help Princeton students find love?

In the end, 646 freshmen, 447 sophomores, 525 juniors, and 637 seniors took the Princeton Marriage Pact, nearly half of undergraduates at Princeton.

“In other schools, not everyone ended up having anything romantic with their Marriage Pact matches,” Park said. “Rather, some of them became best friends because they were so compatible as people. This survey can’t tell you whether there’s that spark that makes something a relationship.”

“We worked so hard at trying to make sure every student’s voice is heard, and even though there are a lot of problems in rolling out Marriage Pact, and [with] the questionnaire itself, I think we did everything we could at every stage to make sure those concerns were addressed immediately, and I hope that at least people came out of this with a good story to tell,” Hyon said.

The Marriage Pact is still constantly being changed and updated, and with every iteration the team behind it continues to strive to improve their company.

“If there’s ever anything that could be done to make the Marriage Pact experience better, like, let’s do it!” McGregor said. “Absolutely. And talk to your friends about it, and send us an email. Whatever it is. Marriage Pact should be a great experience. Whatever it takes, and whatever changes can be made, absolutely.”

Thus, it seems many participants found something of value in the process, and even White is taking his results in good humor.

“[The Marriage Pact is] successful in as much as it’s a service that’s fun or interesting for students to kind of delve into and if there’s just one perfect match on campus where someone meets a significant other, I think that vastly outweighs the weird ordeal that I went through,” White said. “I’m not that broken up about it. I mostly just find it very funny.”