As recently as 2016, the process of delegating trips to students was undertaken entirely by hand — namely, the hands of Director of Community Outreach Amy James, Director of Outdoor Leadership Andrew Jillings, and Assistant Dean of Students for Student Engagement Tessa Chefalo.
“We would print out every registration form on this slow printer in the Glen House and then hunker down in Dwight Lounge with a bunch of tables labeled for each trip,” James recalled. Taxing as this was, the process grew even more complex after orientation trips became mandatory in 2018. “When the number of participants doubled overnight, this method was not going to be sustainable,” Jillings said. “And that was when we turned to the clever professors and the smart kids.”
Enter mathematics professor Sally Cockburn, who — along with a summer research student — created a mathematical model of the situation in response to these increased numbers. “This is a linear optimization problem, and there is software designed to solve these problems,” Cockburn explained. After developing and running their model, the team compared its solution to the one pieced together by James, Jillings, and Chefalo. Showing positive results, this model took over the full responsibility of trip assignments the following year.
This mathematical process is unique to Hamilton. “Everyone else still does the paper thing, or they have a basic first-come-first-serve system,” Jillings said.
Every summer since, Cockburn and a student researcher have tweaked and improved their algorithm, allowing it to account for additional factors in accordance with the purpose of orientation. “It’s a goal of the program to mix students up and let them meet people they haven’t met before,” Cockburn said. Considering these nuances, Jillings said, was “one of the biggest challenges of doing it by hand.” Now, the computer algorithm can account for everything in only a matter of seconds.
This summer, Grisha Hatavets ’25 has been working with Cockburn on further improvements to the orientation algorithm. Aside from adapting the format of the data, Hatavets said, he also altered the model to more accurately account for the number of groups a given number of students would fill. On top of this, Cockburn added, they have been folding in further constraints and details, including gender and sports team variables. “You can test out a lot of scenarios very rapidly,” she said. “When you’re doing it by hand, that’s not really an option.”
Jillings reiterated the value of this adaptability, saying that “in the old days, when we found a solution, we would just collapse in an armchair and be done.” Beyond its convenience, this mathematical process is also unique to Hamilton. “Everyone else still does the paper thing, or they have a basic first-come-first-serve system,” Jillings said.
And the difference is tangible: even in its earliest iteration, the algorithm produced an average placement of 1.6, meaning nearly every student received their first or second-choice trip. At this point, the hardworking orientation team accepted their fate. “The algorithm was just so much better than we were,” Jillings said. “We had to swallow our pride and say yeah, this thing is great.”