Hauber Research Fellowship opens new worlds
August 7, 2017
After spending 10 weeks developing a mathematical model that crunches key National Football League statistics, John Fluck correctly predicted the outcomes of 254 non-tied games from the 2016 season.
That’s a remarkable success rate of 93.3 percent for the 20-year-old Loyola junior—besting nationally-known, professional analysts such as Bill Cower, Jamey Eisenberg, Ron Jaworski, and Chris Simms.
The deeper into the season Fluck went, the stronger the New Jersey native’s statistical model performed.
“The most exciting part for me was just getting the results and comparing them to see how well we stack up with people who do this for a living,” said Fluck, a statistics major. “It’s kind of cool to see that we could match what they’re doing towards the latter half of the season. That was a good feeling.”
Fluck’s project was made possible through the Hauber Research Fellowship program, an annual summer initiative launched in 1988 that pairs undergraduates with faculty mentors from natural and applied science departments at Loyola.
Fellows receive a $5,000 stipend to work on their research topic, devoting 400 hours of laboratory time to their area of interest. At the end of their work, participants submit a written research paper and give a formal presentation in front of other fellows, faculty members and the wider Loyola community. They also participate in a poster symposium in the fall semester and are eligible to be Loyola Student Ambassadors for the University’s annual Cosmos and Creation Conference.
Fluck’s model included key statistics such as completed passes, interceptions, penalty yards, and fourth down conversions. Every other day, the student met with Richard Auer, Ph.D., statistics professor and mentor, to discuss the development of the model and problems that arose in research.
“We planned things out together and he gave me a day to kind of go out on my own and run the model,” said Fluck, noting that a weakness of his model is that it does not easily account for player injuries. “It was really cool that he let me experience the more exciting parts by myself—he called it ‘eating the ice cream.’”
Fluck has no plans to put his model on the line in the gambling arena. But he sees great potential for its applied use in other fields.
“The model we’re using is a really good one to predict events where there are only two possible outcomes,” he explained, noting that it could be used in the medical industry to help predict whether a patient will contract a disease. “How can you use a set of predictors to forecast which outcome you are actually going to get?”
Finding real solutions for real-world problems
Dr. Auer, who has worked with Hauber fellows for a decade, described his mentorship in the program as among the most rewarding experiences of his teaching career.
“You are not in a situation where one is looking down on the other and teaching the other,” he explained. “You are going into it as a pair trying to solve a problem.”
Hauber fellows have the advantage of delving into problems on a level they cannot always experience in the classroom, he said. “They are trying to answer real-world questions,” he explained, “and there’s no answer key in the back of a book. The students gain a ton from this level of hands-on experience.”
Dr. Auer noted that fellows have produced academically solid work that is making an impact in various fields. Well more than half the Hauber fellowships he mentored led to conference presentations, said Auer, who most recently presented a paper at the National Statistics Conference based on work he did with a Hauber fellow in 2012.
“In one case,” he said with pride, “we actually had a publication in a statistical magazine.”
Research topics completed by 2017 Hauber fellows included assessing air quality in urban and green spaces in Baltimore City, examining black tea’s influence on cytokine secretion and investigating the magnetic effects on neutrino masses.
Yon Su Kim, a 20-year-old junior from Maryland, used computers to make the investigation of Medieval Latin texts more accessible. Kim wrote code to search the entire corpus of Latin texts that have been transcribed by Loyola students. Her research dovetails with the work of another Hauber fellow and Loyola junior, Nicole Schneider, who completed a segmentation and analysis of handwritten medieval text for automatic indexing.
“Even though there are millions of Latin words out there, researchers tend to gravitate to the places where we know content can be found,” said Jeffrey Witt, Ph.D., professor of philosophy who mentored Kim and Schneider. Roger Eastman, Ph.D., of the computer science department also mentored the two students throughout their fellowships.
“It’s kind of like being in the dark,” Dr. Witt said. “No one reader can ever read that entire corpus.”
Kim’s research allows documents to be retrieved via a ranking system based on the similarity of certain words to a target document.
“This is a way of letting the computer pull in relevant text from billions of words,” Witt said. “Without machine help, that darkness will always remain unexplored.”
Kim was happy to work on an interdisciplinary project. “I’m not really interested in philosophy,” Kim admitted, “but I’m a computer scientist, and this was a project that took a bit of computer science and philosophy. I’m interested in learning new things, so I decided to explore this topic.”
Working closely with faculty members and seeing the successes and struggles of other Hauber students in the laboratory was rewarding, Kim said. Being a Hauber fellow also opened Kim’s future to the possibility of graduate school, a route taken by many Hauber fellows.
A Jesuit legacy
The Hauber Fellowship program is named in honor of Father Edward S. Hauber, S.J., professor of chemistry at Loyola from 1942 to 1966 who also served as department chair. The priest raised money for chemistry majors to complete summer programs. After his death in 1985, enough additional money was raised to support six students each summer in a variety of disciplines, with additional students supported with grant funds from agencies such as the National Science Foundation and the National Institutes of Health.
Bahram Roughani, Ph.D., associate dean for the Natural and Applied Sciences, said the university’s advancement office has worked tirelessly in establishing an endowed Hauber research account, in addition to pursuing support from industrial partners in recent years including W.R. Grace, Booz Allen Hamilton, Whiting Turner, and Northrop Grumman.
“Our Hauber fellows exhibit confidence when presenting their research work and they reveal a mastery in communicating complex scientific ideas with a diverse group of audience attending the summer research presentations,” Dr. Roughani said.
If a student is thinking of applying, Dr. Witt’s advice is to go for it.
“I’ve never learned anything as well as when I had a project for it,” he said. “It makes the knowledge real in a way classroom exercises don’t. Sometimes you can get exercises done without really knowing what’s going on. Here, you are forced to know what’s going on and why.”