An image from “Digital Diamond Baseball,” a professional baseball simulator created by MCLA’s Dr. Mark Cohen and Dr. Les Morey of Texas A&M University, that models real-life baseball and produces accurate, simulated game play.
Computer Science Prof Keeps Current in Rapidly Changing Industry
There’s no need for Dr. Mark Cohen’s computer science students to ask him what he did over his summer vacation. They’re learning it in his classroom.
This past summer Cohen collaborated with musician Robert Phelps, an artist associate at Williams College, to develop a cell phone application that helps teach people how to play the guitar.
“He provided the guitar teaching expertise, and I provided the technical expertise,” Cohen said.
Called “Guitar Quest,” the game teaches users to play the guitar as they solve “mini quests” to achieve various goals. “The motivation was for me to have a real project that would allow me to learn a lot more about cell phone app development,” Cohen explained.
“The work was done in the summer in preparation for redesigning our senior-level project class in the fall,” he continued, “so that students could learn cutting-edge, cell phone application development technologies. Everything covered in that class was a result of what I learned from this project.”
According to Cohen, there’s never a moment’s rest in the rapidly changing computer science industry.
“Every summer I go out and learn new stuff. Then I can redo, essentially, the content of my courses. The courses stay the same, but what we teach is different because there are constantly new things,” he said.
Six years ago, Cohen created a software development company called Curiosity Computing. In addition to the “Guitar Quest” app, he offers a professional baseball simulator, “Digital Diamond Baseball,” which models real-life baseball, and produces accurate, simulated game play.
The computerized baseball simulator, which Cohen created with Dr. Leslie C. Morey, a member of the department of psychology at Texas A&M University, allows gamers to play individual games, series or entire seasons, using players from any baseball season in history.
In addition to being a commercial pursuit, the baseball simulator, found at http://digitaldiamondbaseball.com/,also is a research project which resulted in the recent publication of a manuscript Cohen co-authored with Morey.
Their article, “Bias in the log5 estimation of outcome of batter/pitcher match-ups, and an alternative,” was published in the Journal of Sports Analytics at the end of 2015.
It’s basically about the new algorithm for predicting matchups between pitchers and batters, and evidence that the algorithm was working. That evidence was derived from many, many runs of simulations using the game to gather data,” Cohen explained.
The audience of that manuscript is actually baseball executives,” he continued, “so major league baseball teams and scouts and so on definitely have an interest in being able to predict, at the probability level, the likelihood of a certain outcome when a batter and a pitcher meet up.”
Cohen said the paper has caused quite a stir.
“We’ve gotten a lot of feedback from people about that algorithm and its usefulness. We’ve been contacted by leaders in the field of sabermetrics, who agreed and disagreed with our results, so it’s created some interesting academic discussions.”
Projects like these not only allow Cohen to learn more about his field in an interesting way; they help him stay current in the computer science industry as he conducts research and explores new technologies to bring into his classrooms.
“I definitely love to teach and I like to learn,” he said. “These games give me the chance to learn new stuff, get excited about it, and then share the technologies I’m learning with the students.
“It’s a blast. I absolutely love it,” Cohen added. “It’s why I went into teaching.”