UMF scholar trying to line up optimal batting order

0

FARMINGTON — Sean Cabaniss spends his days doing what rabid fans of America’s pastime dream of: boiling down pro baseball lineups into mathematical formulas.

The University of Maine at Farmington fifth-year student and Wilson Scholarship recipient is on a quest to discover an optimal Major League Baseball lineup through a research project.

“I wouldn’t be researching baseball if I didn’t love the sport so much,” said Cabaniss, a diehard Boston Red Sox fan. “I’ve played it for 18, 19 years of my life.

“It is certainly near and dear to me, but I never really thought I would do this much research with baseball, but it is also an awesome way to combine athletics and academics and just the research side of it.”

Advertisement

Cabaniss, who majored in secondary-education mathematics and played baseball for four years at UMF, grew up in Belgrade and resides in Vassalboro.

“Math plays a role in everything. My love for math — I see it everywhere,” he said.

But he isn’t taking a breather from athletics. He had eligibility left to participate in the Beavers’ track program as a jumper and sprinter. He helped with workouts for the ski team on the men’s side, too.

For an encore, the math whiz is compiling mounds of baseball data and will feed the information into a computer.

“It’s a great opportunity to work on an undergraduate research project that combines mathematics, computer science and baseball,” said Chris Bennett, an associate professor of computer science at UMF. “The baseball aspect certainly makes it more approachable for a wider audience while still offering a chance to do some more advanced mathematical and computer science work.

“(Cabaniss has) been committed to the work, and it is already showing some interesting results,” Bennett said. “Like any research project, there’s a chance the research uncovers something truly groundbreaking, but this is far more often not the case; these sorts of “discoveries” require not only hard work but a lot of luck.

“Perhaps the biggest significance is the exposure for Sean to the research process: identifying hypotheses, creating experiments and exploring proofs, synthesizing and disseminating results. These are the kinds of skills that will serve a student in a career after college.”

How does this work?

The process of reaching a conclusion requires mathematics and computer simulations — and there are no guarantees Cabaniss will find an answer.

“So what we are doing is taking a look at the nine positions in the batting lineup and we are trying to see which permutation is going to pretty much optimize the batting lineup,” Cabaniss said. “Permutation meaning the amount of ways you can rearrange the lineup. So because there are nine positions in the lineup, there are nine factorial ways you can arrange that or over 360,000 different ways that you can arrange a baseball batting lineup.

“So we are trying to find the optimal way to do that, and through doing that, we are taking a look at people’s stats. We are going to see how often a player or position in the lineup comes up in a certain situation.

“We are running computer simulations right now to see actual lineups’ optimization. We have a formula that we can plug players’ stats into and it will output a number that will tell us their effectiveness. This formula takes into account plate singles, doubles, triples, home runs, productive outs, unproductive outs and double plays.”

Out of all this research will come formulas for the optimal lineup that could provide maximum output on offense.

“Yeah, it’s got coefficients that we use to determine how valuable a single, double is, a triple is, home run, walks and that kind of thing and we will put in a player’s actual stats in an actual lineup,” Cabaniss said. “Most recently, I looked at the American League championship series teams from last year and looked at National League championship teams and put together lineups.”

The inquisitive UMF student, along with professors Bennett and Nicholas Koban, have teamed up to work on the project.

“We can run thousands of games in seconds through a computer,” Cabaniss said. “So what we are really trying to see is how productive a certain player can be in a certain position of the lineup.”

Bennett is an advocate of Cabaniss’ undertaking and the project.

“At the least, a research presentation describing the project and its results will come out of it,” Bennett said. “And perhaps something really interesting is uncovered along the way.

“From the undergraduate student’s perspective, the project is difficult enough to need mentoring from professors,” he said.

“Another unique aspect that affects its difficulty is that it brings together three fields: mathematics, computer science and baseball. Some understanding of all three is required to do a project like this, and Sean’s in a unique position in this way,” Bennett said.

Picking up the trail

UMF student Tucker Atwood started the research for the optimal baseball lineup.

“He worked with Dr. Nicholas Koban and also Dr. Chris Bennett and they made some leads on it, but we took a different approach this year,” Cabaniss said. “I approached Nick and I wanted to do some more research on it and we decided this was something that would be really cool to continue, take some different avenues. So he proposed it to me.

During his work, Cabaniss discovered many ways to “permute a lineup.”

“If you look historically at baseball, the average number of runs per game is four,” he said. “You are going to have games where you don’t score any or put up 11 runs.

“But there is a possibility of looking at this equation and entering in certain players that could produce more than four runs per game and see where that goes,” he said.

He also pointed out that an answer will be difficult to ascertain because of a number of factors that influence a lineup in each game.

“It’s going to be hard to tell because a lineup might change from day to day based on the pitcher you are facing,” he added. “There is a lot more that goes into it, too.

Cabaniss remains undaunted, and for him, the journey and self-discovery is just as important as the conclusion.

“If we don’t get a conclusion, it can be continued by somebody else,” he said. “I will be presenting a symposium, which is an annual symposium at UMF, which all the Wilson research scholars have to present.”

Another door opens

The research project opened Cabaniss’ eyes to other opportunities besides a career in teaching. After he leaves Farmington, he will attend Smith College in western Massachusetts, where he has been accepted into a master’s program for exercise and sports studies. He believes a master’s degree will lead to a college coaching career

“Originally I wanted to teach and I think that still might be an option for me down the road,” Cabaniss said. “The master’s program I am heading into now is really setting me up to be a full-time college coach at some point — a college track coach — even with all the math.”

University of Maine at Farmington fifth-year student and Wilson scholar Sean Cabaniss is using mathematical formulas and computer simulations to come up with the opitmal Major League Baseball lineup.

Sean Canabiss gets ready to hit a pitch during a University of Maine at Farmington baseball game last year.

Advertisement