Keeping the Bar High, New Jersey High School Fields Top Two Teams in Modeling Contest

June 12, 2007

Manalapan students not only have great math and science skills, say coaches Jessy Friedman (left) and Stephanie Pepper; in addition, “they can interpret, organize, and write programs for data related to any type of problem. . . . They all work in teams on their senior projects so collaborative effort is not new to them.”

On April 18, for the second year in a row, SIAM joined The Moody's Foundation in New York City for the final stage of the Moody's Mega Math Challenge, followed by a reception for the top six teams. The M3 Challenge is a Web-based mathematical modeling contest for metropolitan New York City high school students, who are given an open-ended, realistic mathematical modeling problem and 14 hours to work on it, availing themselves of any nonhuman resources they can find. Moody's provides financial support, including scholarship awards (with a first prize of $20,000 each year, and awards totaling $65,000); SIAM administers the event.

The enthusiasm and creativity of the students would be story enough, but this year's challenge had the added attraction of a single school---Manalapan High School, a public science-and-engineering magnet school in New Jersey---that took the top two prizes. Some of the students, and both coaches, had participated in last year's challenge as well. Interest in the contest is intense, says Stephanie Pepper, one of the coaches: "We had to draw names from a hat to select 10 seniors to make up the two teams."

Interest in the contest was high in general, with numbers increasing from 129 teams (572 students) in 2006 to more than 156 teams (700 students) this year. The contest also attracted media attention, not only local newspapers but also CNBC, which had a crew on the scene on April 18, filming student presentations and then the award ceremony in the evening; in the end, a substantial segment on the event aired on national television.

This year's problem, consistent with the downtown location of the M3 Challenge sponsor, was titled "Beat the Street!" The students were to develop and test a model for building an investment portfolio, maximizing net profit. With $30,000 to invest and 18 stocks to choose from, the students were to purchase any combination of up to six stocks, selling them after one year.

How did two teams from the same school emerge on top (with, according to judge Jim Crowley, solutions reached by very different approaches)? Coaches Jessy Friedman and Stephanie Pepper (of this year's first- and second-place teams, and last year's eighth- and fifth-place teams, respectively) attribute the stellar performance of the Manalapan teams in part to the nature of the school. "The kids have a very rigorous schedule for all 4 years," they wrote in response to an e-mail message from SIAM News; they spend their senior year "doing research and presenting projects that cover the spectrum of engineering and technology." An open-ended modeling problem "is precisely what their senior projects are all about."

The Manalapan students are friends who in four years have become "extremely comfortable with each other, . . . yet with a healthy level of friendly competition," Friedman and Pepper wrote. The teams may derive an edge, they continued, from the combination of courses they take: AP courses in the sciences and high-level math, in addition to statistics and computer science.

Teams of judges evaluated submitted solutions in stages. For the final stage, on April 18, four judges---Jim Crowley, Ben Fusaro, Henry Ricardo, Lee Seitelman---
heard presentations by the six highest ranked teams.

The problem was intended to allow students who had taken standard math courses to participate, while permitting more sophisticated approaches as well, Crowley says. "Teams approached the problem in a variety of ways---statistical approaches, Monte Carlo simulations, even Markowitz valuation theory. Solutions were accompanied by impressive graphs and even, in some cases, Java code."

"Teams are given only 14 hours to model a problem and analyze its results," Crowley explained at the reception. "They do not know in advance what the problem is."

"This is not atypical of what a mathematician working in industry must do," he pointed out. "Although they do have some knowledge of the area of the problem, an industrial mathematician is often asked to quickly develop a mathematical model and then use it to analyze a situation."

From a student perspective, that uncertainty is a big part of the challenge. The first year's problem, which entailed a restructuring of the U.S. Social Security system in a way that would keep it solvent well into the retirement years of the modelers, surprised students who had expected a "more traditional" applied math problem. (Sample problems and other materials are posted on the Web well before the actual contest day; This year, many competitors were surprised anew: "I was caught a little off guard by the topic," says Naiim Ali, a member of this year's first-place and last year's fifth-place teams, "but once we started working on the problem I started to like it more."

Many of the students seemed to share the sentiment. In fact, Crowley reports that several teams followed their stock picks during the month or so between submitting their solutions and presenting them to the judges. "Several students made a point of telling us how well their portfolios were doing," he says.

Renew SIAM · Contact Us · Site Map · Join SIAM · My Account
Facebook Twitter Youtube