# For This Contest, a Handle on Statistics Helps Forecast the Future

How many hits will Taylor Swift have on the Top 40? Competitors crunch the numbers for this and other questions

When the New York Mets’ DJ Stewart hit a powerful three-run homer in the bottom of the sixth at a home game against the Chicago Cubs on April 30, it likely made some aspiring statisticians very happy.

Statistics have always been a big part of baseball. But in this case, the outcome of that game—which the Mets won 4-2—was integral to a contest from the American Statistical Association, devised by three faculty members from Tufts University School of Dental Medicine. (Yes, you read that right. We’ll get to the dental school angle in a minute.)

The contest asked high school, undergraduate, and graduate students across the world to predict the outcome of five events—ranging from sports to pop music to the stock market—taking place during the month of April, which is National Mathematics and Statistics Awareness Month.

Those without a good grounding in the field might be puzzled how statistics can be used to help predict the number of victories for a baseball team, much less one that’s been known to disappoint its fans. Or, for that matter, to predict how many tunes Taylor Swift will have in the Top 40, or what the closing stock price on the Nasdaq for Intel Corporation will be. But all of those were among the questions presented in the Predict 5 contest.

“A lot of statistics is using the data available to you, and extrapolating from that,” says Assistant Professor Shruti Jain, who, along with Associate Professor Matthew Finkelman and Assistant Professor Sarah Pagni, created the contest.

Each prediction offered multiple potential approaches, says Finkelman, who is the director of the dental school’s Biostatistics and Experimental Design team.

“There are a lot of different ways people can go, from the simple to the sophisticated,” Finkelman says. The simplest way in the baseball question, he says, might be to assume that the Mets will win half their games this April. Or you could take into account the number of games they won last April. Or, still more complex, how the roster has changed; their opponents; how many games will be at home; the dimensions of the ballparks they will play in; weather forecasts—the possibilities are endless.

This process is similar to what scientists do when they are designing and conducting experiments, Pagni says. Both involve determining what data to use, analyzing it, and reaching a conclusion.

“We take data from a sample and use that data to make a conclusion about a larger subset, or an overall population, and then determine is that actually the case, or was it something that was just random chance?” she says.

At TUSDM, Finkelman, Pagni, and Jain collaborate on research projects with faculty and students.  “We’re involved in the life cycle of the research, starting from the conception of the idea,” says Jain.

They also teach statistics and experimental design to predoctoral and postdoctoral students. It was a couple of those first- and second-year courses that captured the interest of Jonathan Im, D26, who was among the winners of the Predict 5 contest, which was judged on which contestants came closest in their predictions.

“I don’t come from a statistical background,” Im says. When he first arrived at TUSDM, he would have been surprised to learn he’d one day enter a statistics contest, and receive an honorable mention. While he still prefers the patient-facing side of dentistry, he’s become part of a student research team, collaborating with Jain, that presented in March at the dental school’s annual research fair, known as Bates-Andrews Day.

The team examined data gathered during previous TUSDM service trips to Zambia. “We were able to draw conclusions that will help us in future trips,” Im says. “And maybe this research can help in making dental clinics in Zambia better, more efficient, better for the people of Zambia.”

The impact of research conducted at TUSDM—whether it’s about public health, materials used in dental offices, or treatments for oral diseases—is extensive, Pagni says. “What we’re doing now will affect people’s mouths in the future.”  Finkelman agrees. “Biostatistics is really meaningful. We have an opportunity to have a positive societal impact, and that’s really important to me.”

But before he discovered biostatistics and oral health research, there was baseball.

“As a kid,” Finkelman says, “ ‘How many games will the Mets win in April’ was the kind of thing I was really curious about. And then it kind of expanded from there.”