It may still feel like the dead of winter in many places in the U.S., but baseball’s more than 500 million American fans are already looking forward to spring and their first forays to their local ballparks. While players ramp up their off-season workouts in preparation for the 2023 season, a team of computer scientists and mathematicians is working behind the scenes on one of the sport’s most challenging problems: scheduling games for professional teams.
Just ask Nick Lu. A third year double major in major and applied mathematics and statistics and computer science, he is part of a team of Hopkins students tackling the creation of a schedule for members of the Frontier League, a professional independent baseball league with 16 teams in the northeastern and midwestern parts of the United States, well as eastern Canada.
Lu became interested in baseball scheduling after conversations and classes with Donniell Fishkind, an associate research professor in the Department of Applied Mathematics and Statistics who has been working with students on baseball scheduling for almost 10 years.
In Fishkind’s Intro to Optimization I and II courses, Lu discovered what he describes as the “beauty of organizing formulas to fit specific criteria,” and was eager to be part of the scheduling effort. The team is supervised by Fishkind, Anton Dahbura, a professor in the Department of Computer Science, and alumnus Eli Katz ’20, a data scientist at Big League Advantage and project manager of the Johns Hopkins Baseball Scheduling Research Team. Katz relates to students like Lu because he was a member of the JHU Baseball Scheduling Team beginning in his sophomore year at Johns Hopkins. He now helps recruit and train new student schedulers.
“Baseball scheduling is such an interesting puzzle to solve, both mathematically and practically when it comes to the desires of each league. I want to share that enthusiasm with the students and get them involved in an enriching and hands on operations research experience,” said Katz.
Lu appreciates the opportunity to work with Katz, who understands the scheduling team’s experience from both a professional and student perspective.
“One thing I learned from Eli Katz is the ability to identify where the code may go wrong and know where to fix the problem if the result is not satisfactory,” said Lu.
Lu and his five-person team use MATLAB, a programming language that allows for the plotting of many different functions and data—as well as the implementation of algorithms—to enter variations as mathematical expressions to represent specific constraints imposed by the professional independent baseball league.
Each baseball team submits a list of requests to the league, which may include anything from playing on their home field on Father’s Day to hosting a preferred visiting team at home for the opening series. In addition to implementing various requests, the scheduling team ensures that the baseball team’s schedule has a close-to-equal balance of home and away games.
Maintaining equilibrium while fulfilling multiple requests at once has been a formidable challenge.
Balancing the relative importance of each request makes fulfilling numerous requests at once a challenge, Lu said.
Lu said that the Frontier league presented more than 20 constraints (priorities such as various dates, locations, intervals, and other special requests), and for each constraint, three loops, a sequence of instructions that is continually repeated until a certain condition is reached, were coded to represent a specific requirement. He said that it usually takes up to three days to optimize a schedule running the requirements using MATLAB on a supercomputer. However, the team often received additional league requirements, potentially prolonging the process.
Finally, after almost a month of work, Lu’s team submitted eight options to the Frontier League, which approved the team’s 7th submission. Even so, the team plans to continue working on the league’s schedule, analyzing the advantages and costs of changing the number of requests permitted for each baseball team.
“My first experience working with the Hopkins Baseball Scheduling Research Team was quite enjoyable, and I am looking forward to continuing to work on scheduling for a variety of other leagues in the future,” said Lu.