A classic cautionary tale against binge drinking in college paints a picture of a student who consumed far more than his tolerance level and slipped into unconsciousness at a party, slowly dying from alcohol poisoning as his unaware classmates draw crude images on his face with marker. But some researchers want to use math to better understand how college students drink in hopes of preventing a scenario like that from happening.

The team, in an article from The Ohio State University, compared the way college students maintain their buzz to the way cruise control works on a car: College students in a study “drank until they attained a certain level of drunkenness, and then adjusted the pace of their drinking — sipping versus gulping, for example, or switching to a non-alcoholic beverage — at different times throughout the night to maintain that level.” Now they want to analyze the study data, including blood-alcohol levels and student questionnaires, “via engineering methods” to devise more effective prevention.

Researchers are also collecting more precise data, putting alcohol monitors on 60 OSU seniors to record blood-alcohol data through sweat, similar to the kind used in the criminal justice system, as well as personal fitness monitors to measure sleep and exercise. During the study period, the students will fill out surveys on their phones with “questions about their health and well-being, such as their emotional state at that moment.”

The projects are seeking to understand how the different factors that influence drinking behavior work together. “We’re looking for the best points to intervene strategically, so that we can aid a person in their decision-making and potentially derail problematic behaviors,” John Clapp, director of OSU’s Higher Education Center for Alcohol and Drug Abuse Prevention and Recovery, said in a university statement.

Among the many other impacts alcohol has on society, almost 88,000 people suffer alcohol-related deaths in the United States every year, according to the National Institutes of Health. That could mean anything from a drunk driving fatality to alcohol poisoning. In the case of college students specifically, more than 1,800 die from car crashes or other unintentional injuries, and almost 800,000 are either assaulted by a drunk peer or experience alcohol-related rape or sexual assault.

“We could track as many as 5,000 different variables per person during that two-week period, plus all the social interactions between the people in the different groups,” Clapp said in the statement. “We’re hoping to get a very rich, complex dataset, and most social science methods wouldn’t lend themselves well to untangling all of that.”

From the engineering perspective, Clapp’s partner and OSU engineer Kevin Passino wants to look for “controllability,” or ways to change the system presented. “You can test for controllability with mathematics, and we’re hoping that doing so will suggest some ideas for an intervention,” he said in the university statement.

According to OSU, the goal is having information to develop a mobile app that will alert someone when they’ve had enough alcohol, help them control absorption by suggesting food or non-alcoholic beverages at certain times, and help call a cab to reduce drunk driving.

An app to assist drunk people using math — or other methods — is not a new idea, but so far has been largely limited to the social aspect of being drunk. One app aims to stop the dangerous practice of calling someone in the contacts list while inebriated by locking the phone with math problems, and another helps partying people locate one another.