The New Gen Ed Lottery System, Explained
Armed Individuals Sighted in Harvard Square Arraigned
Harvard Students Form Coalition Supporting Slave Photo Lawsuit's Demands
Police Apprehend Armed Man and Woman in Central Square
107 Faculty Called for Review of Tenure Procedures in Letter to Dean Gay
Most people would not spend days strategizing how to win a foosball game, but in this year’s second annual Student Computational Challenge, teams were challenged to do just that and code a program to trump opponents in a virtual tournament.
James A. Damore, a second-year Ph.D. student in systems biology, and Bo Waggoner, a second-year Ph.D. student in computer science, beat seven other student teams Wednesday during the event, winning a pair of iPad Minis as reward.
“In the end, our solution was simple, but we were actually pretty confident it was optimal,” Damore said.
The challenge, which opened Monday morning in Maxwell Dworkin, required the teams to submit algorithms which then faced off in a round robin tournament. In the winners’ final implementation, Damore said he coded most of the algorithms while Waggoner designed methods to test and visualize their performance.
Pavlos Protopapas, a researcher at the Harvard School of Engineering and Applied Science’s Institute for Applied Computational Science, helped organize the competition as a part of ComputeFest—a week of Wintersession activities planned by the organization.
While last year’s challenge was an optimization problem for teams to solve on their own, Protopapas decided that a sports game would provide a more interactive experience for this year’s competitors.
“I wanted the challenge to be a little more dynamic, a little more fun,” Protopapas said.
In the game, students submitted a starting lineup of players, distributed as they wished across the virtual field. As the ball moved across the field, a server communicated that information to the students’ programs, which they could then use to reposition the players.
Damore and Waggoner’s strategy was to first simplify the problem, imagining a case where they could not reposition players. They found that the best solution was to place the most players in the midfield and then adapt that strategy to the more complicated problem.
“We had to test everything against different strategies,” Damore said. “It would be very easy to be blind to certain strategies and then lose to those in the competition.”
Protopapas said he designed the challenge so that it could be approached using a variety of methods. At the end of the competition, the top four teams presented and discussed their individual strategies.
—Staff writer Brian C. Zhang can be reached at firstname.lastname@example.org.
Want to keep up with breaking news? Subscribe to our email newsletter.