NEWS & EVENTS

Dordt College News

Math prof lends hand to youth hockey

January 15, 2010

When mathematics professor Gary De Young sees an opportunity to use his talents and make a difference, he tends to jump in and see how he can help.

Gary De YoungDe Young moved to Sioux Center in 2006, accompanied by two sons who play hockey.

They quickly became involved with the local youth hockey program, and De Young quickly became aware of the complexity of annually scheduling nearly 600 games at ten rinks, each with its own rink schedule.

“There are numerous scheduling conditions, making this a big problem,” says De Young.

Prior to this year, scheduling was done by hand, which introduced many errors. It was also very time consuming to check. Ten local schedulers spent hours coordinating changes and adjustments via e-mail.

De Young had two goals:

a) Provide online tools to collect, disseminate, and manipulate schedule information.

b) Provide an initial schedule that incorporated known scheduling constraints like the number of games possible at a rink, referee availability, spacing of games with opponents, the number of venues used on a given day, consecutive travel days, etc.

To achieve his first goal De Young set up a website and database to collect and display information in a user friendly way, and he created various tools for manipulating and analyzing schedules. The second goal involved more mathematics. Using sets and functions, he represented possible schedules as sets and developed a function that modeled the “goodness” of the set (schedule). De Young’s task was to find a set that optimized the function.

Nearly fifty-five percent of the games on this year’s initial schedule remained unchanged

“One local scheduler said this was far fewer than in the past,” said De Young, who learned a great deal about scheduling preferences in the process. He hopes to work what he learned into next year’s optimization function for the initial schedule.

“Maybe next year I’ll have seventy-five percent remain unchanged,” he says. “Modeling scheduling preferences with function optimization is an art, not a science.”


SALLY JONGSMA

Media Access: Download Word Version | High Resolution Image: 1