A referee holds back a football player on the field.

Students’ study finds bias in football rulings on the field

Mike Macey, Rhett Brymer, and Mickey Whitford, gather around the desk in Brymer's Farmer School of Business office.
Mike Macey, Rhett Brymer, and Mickey Whitford analyzed football referee play calls looking for incidents of potential bias.

When you’re a frustrated football fan, one of the phrases you probably utter the most is, “Come on, ref!” Rhett Brymer, the John Mee Endowed Assistant Professor of Management, and undergraduate students Mickey Whitford and Mike Macey are no different — except for the fact that they have access to big data analytics tools and the skills to use them.

In 2013, Brymer decided that sitting on the sidelines as a football fan wasn’t going to cut it, so he set out to find what, if any, biases existed in college football officiating. He started sorting through mountains of footage from 2005 to 2012. This study was presented at the MIT Sloan Sports Analytics Conference, and Bloomberg contributor Bryan Gruley picked up the story.

It turns out there’s more than one way to determine bias in referees. The traditional way, which is used by the NCAA officiating teams, is for monitors to view game film and judge how good or bad certain calls are and see if there are any systemic biases, such as referees throwing games or tournaments. The second way, which Brymer ascribed to, was to use data analysis to make the same determinations.

Most recently, he began to collect data on the years from 2012 to 2015. This came out to more than 38,000 individual observations put into one Excel sheet. Each observation represented an instance where a foul was called on the field. As you can probably imagine, this was a huge amount of data.

Here’s where Whitford and Macey got involved. They took a data analysis course in spring 2015 with Waldyn Martinez, assistant professor of information systems and analytics. Martinez mentioned Brymer’s research as a way that data analytics can be applied to the real world, and the students’ ears perked up.

Macey (a senior accountancy and analytics double major) had just told Whitford (a senior supply chain and operations management and analytics double major) the day before that he wanted to do some sort of analytics project, and Whitford had just thrown around the idea of doing something in sports analytics. So, this project seemed too serendipitous to pass up. They got in touch with Brymer, and the rest, as they say, was history.

Finding the right model

The students’ main task was to take the data — the incredible 38,000+ observations —and turn it into something readable. The analysis itself took them close to nine months.

“Mickey and Mike did the heavy lifting,” Brymer said. “They’re doing this for no class credit, no pay,” he noted.

“Once we got the finalized model, it was all about making it more interpretable and including the right parameters, such as year or performance,” Whitford said. “We did a lot of the computation and data analysis, and then Dr. Brymer wrote the article and made sense out of the numbers.”

One of the biggest findings of Brymer and the students was that on average, referees are ten percent less likely to throw discretionary flags (for fouls like holding, unsportsmanlike conduct, pass interference or personal fouls) on teams more likely to go to the playoffs or that have winning traditions. Brymer, Whitford and Macey found that these teams, which they called “protected flagships,” did especially well in the Big Ten conference.


Written by Elizabeth Jenike, Farmer School of Business, Miami University. Originally appeared as a “Top Story” on Miami University’s News and Events website.

Photo of Mike Macey, Rhett Brymer, and Mickey Whitford courtesy of the Farmer School of Business, Miami University. Referee and player photo by AJ Gruel via Wikimedia Commons, used under Creative Commons license.

 

A red brick building with cream trim is visible between two trees with fall foliage. The building is four or five stories tall, with big columns in the front and a cupola on top. Several people are walking on the sidewalk in front of the building. Green manicured lawns are also visible in front of the building.

Undergraduate students present winning paper at national conference

Two young women hold a glass award. Visible on the award are the words "Best Paper" and "UCIS." Between the two women stands a man who is older than they are. All are smiling.
Miami University senior Rachel Frame (left) and Sydney Powell (right) presented a paper they co-authored with Dr. Jeff Merhout (center) at the Undergraduate Conference in Information Systems.

Farmer School of Business seniors Rachel Frame and Sydney Powell were invited to present the research paper they co-authored with Dr. Jeff Merhout, associate professor of information systems and analytics,  at the prestigious Undergraduate Conference in Information Systems hosted by Carnegie Mellon University. This research was completed to meet the University Honors Undergraduate Research Forum requirements for Frame and Powell.

Their paper, “Understanding the lack of women in IT/IS,” tackled an issue close to their hearts, and important to both educators and employers. Their research examined the impact that stereotypical gender roles and exposure to the information technology (IT) and information systems (IS) fields can have on both male and female perspectives towards the IT/IS field. Results of their survey indicated that females are either less experienced in, less exposed to, or less knowledgeable than males about IT/IS, which likely leads them to be less interested in and therefore lack a presence in IT/IS fields.

Their research, writing and presenting skills garnered them the Top Paper Award at the conference. “I am very proud of the outstanding research and write-up that Rachel and Sydney completed under my guidance. This work truly makes a contribution to our understanding of why Miami students do or do not choose IS as a major,” noted Merhout.

Written by Addie Rosenthal, Senior Director of Communications, Farmer School of Business, Miami University.  Originally appeared on the Farmer School of Business website.

Photographs by Miami University Communications & Marketing.

Photograph of the Alcoa Corporate Center in Pittsburgh. The building is made of glass and steel and is sort of wave-shaped. The blue sky and white wispy clouds are reflected in the building's windows.

Students’ analysis of data helps international company find path to reducing employee turnover

Words related to human capital are arranged to form the head-and-shoulders silhouette of a person in profile. The words include "social," "competence," "management," "work," "people," "job," "manager," "recruit," "theory," "leader," and "development." At the bottom of the silhouette, the words "Human Capital" appear in larger type of a different color so they stand out as a title for the image.
Miami University management professor Dr. Josh Schwarz teaches a class on human capital metrics. Last semester students in the class worked on a real-life problem experienced by client Alcoa.

You can’t manage what you don’t measure. This adage applies to everything in business from customer loyalty and retention to overhead costs. According to Miami University management professor Joshua Schwarz, it even applies to human capital.

“Using data to provide insights into management is something that all organizations need to be doing,” Schwarz says. “So many people manage based on gut feeling, but evidence-based management – using good data to make managerial decisions – is more effective.”

Schwarz teaches a course on human capital metrics in Miami’s Farmer School of Business. Students in this course collect and analyze data relevant to current business problems facing real clients.

“The class is about quantifying various aspects of people management and showing connections to measures of organizational performance,” Schwarz says. “Students work on a broad range of problems. The thing that unifies them all is that they’re all about data analysis.”

Since 2005, Schwarz’s students have completed more than 150 projects for dozens of clients of all sizes, including GE Aviation, Cleveland’s Rock and Roll Hall of Fame, and the City of Dayton.

This past fall, Schwarz’s class worked with Alcoa, a global leader in lightweight metals technology, engineering and manufacturing, to help reduce employee turnover at one of its locations. Turnover is a problem for many businesses because hiring and training replacement workers is expensive. According to a 2012 study by the Center for American Progress, those costs, plus lost productivity, can total as much as 20% of the typical exiting employee’s annual salary.

Alcoa wanted to know if a three-hour-long pre-employment assessment they were already administering could help predict which job applicants had staying power. So they came to Schwarz’s class for help developing a candidate profile that correlated performance on the assessment’s 12 dimensions to employment longevity.

Five of Schwarz’s students took on the project: senior management and leadership majors Griffin McCarty, Alexandra Ohly, and Alexa Blumling (who graduated in December), senior psychology major Marisa Murray, and junior psychology major Deepika Hebbalalu. Both aspiring human resources professionals, Hebbalalu and Murray were especially eager to work on the project.

“This was my first consulting project,” Hebbalalu says, “and it was an amazing experience to work on an actual business problem and apply what you learned in class so directly. As a psychology major, I learned all these statistical analyses and methods in class. It was exciting to actually get to apply them to a real data set. It made it really relevant.”

Murray agrees, and adds, “To have the first professional experience of talking with a client was really cool too.”

Hebbalalu, Murray, and the rest of their team compared employees’ total and individual dimension scores to how long they stayed with Alcoa after being hired. Ultimately, they were unable to find a correlation between Alcoa’s screening test, as a whole, and which job candidates were likely to stay with the company.

“However,” Murray says, “we found that the reading and cognitive dimensions were correlated to whether an employee stayed with the company.”

For Alcoa, knowing what doesn’t work may be just as valuable as knowing what does work. That’s because they stand to reduce costs by following the team’s recommendation to eliminate the ten non-predictive dimensions, and give job applicants only the reading and cognitive dimensions going forward.

“Shortening the test would require less time from a proctor and fewer test materials,” Hebbalalu says, “so it would be cheaper to administer. We did a calculation of return on investment if they were to revamp the test to figure out how much money they’d save.”

Schwarz cites the team’s work as a straightforward example of the value his students offer their clients. “In this case,” Schwarz says, “the students were able to clearly demonstrate return on investment because all they were proposing to do was cut down on costs so that the company wasn’t spending money on something that wasn’t adding value.”

Having completed Schwarz’s course, Hebbalalu and Murray now serve as undergraduate assistants in the spring semester class, where many of the projects focus on recruiting and retaining Millennials. It’s a way for them to give back and pay forward at the same time. Not only are they using their experience to help current students navigate the course, Hebbalalu and Murray are also helping their fellow Millennials shape the workplace they will soon enter, and even manage.

Written by Heather Beattey Johnston, Associate Director & Information Coordinator, Office for the Advancement of Research & Scholarship, Miami University.

Photograph of Alcoa Corporate Center by Nick Foust via Flickr.  Used under Creative Commons license.