Now in its third year, the Wharton High School Data Science Competition – hosted by the Wharton Sports Analytics and Business Initiative (WSABI) and Wharton Global Youth Program – brings together students from around the world to tackle complex problems at the intersection of sports and analytics. The competition this year, which is presented by Google Gemini, saw more than 700 teams from 48 countries analyzing simulated hockey data from the “World Hockey League” to uncover meaningful insights about the sport.

After a few rounds of eliminations, one team remained: “2016 Chino Hills” from Pingry School in Basking Ridge, New Jersey, named after the legendary high school basketball team, known for their aggressive offense and improbable 35-0 record.

This iteration of 2016 Chino Hills was comprised of juniors Zach Z., Max L., Lucas G., Aaron W., and Riley W. The team didn’t begin as a formal unit. After participating in Wharton’s Moneyball FLEX program in the summer of 2025, Zach returned to Pingry energized to go further. “I was looking around the website for more ways to get involved in sports analytics and came across the competition,” he said.

He soon founded a sports analytics and data science club at Pingry, bringing together classmates with a range of interests, from baseball and football to swimming and motorsports. With Aaron helping recruit members and Lucas stepping into a technical leadership role, the group quickly grew. From that larger pool, the eventual winning team took shape.

Building a Model—and Learning Its Limits

Like many competitors, the Pingry team entered the competition drawn by the opportunity to apply technical skills in a real-world context. “I really enjoy the numbers behind sports,” Zach explained, “and I wanted to find a way to use that.”

Their approach to crunching the hockey data centered on building a Bayesian model, a method that allowed them to uncover patterns and probabilities within the dataset. Using R and a combination of AI tools, the team worked through multiple iterations, carefully validating outputs and cross-checking results.

At one point, everything seemed to click. “When the Bayesian model came together and produced the output we were aiming for, that was my favorite part,” Zach said. “We sanity checked all 12 models, and all the numbers made sense.”

What initially presented itself as progress, though, soon turned out to be the beginning of a setback. After completing much of their analysis, the team realized they had overlooked a key variable: home-ice advantage. Accounting for it meant revisiting, and ultimately rebuilding, their model. “It made it so we had to go back and redo our entire Bayesian model,” Zach recalled.

As Lucas put it, “Data analysis is not a linear process… when you have to start from scratch, it can be demoralizing, but it is important to stay on track and stay positive.”

Long Nights, Teamwork, and Different Ways of Thinking

Behind the final model was a workflow shaped by busy schedules and shared commitment. Over the course of the competition’s two-month schedule, the team balanced group sessions with independent work, often coming together to refine ideas and troubleshoot challenges.

“I do remember a few nights working straight through from 12–7 a.m.,” Zach said. “To be honest, it didn’t really feel that long once I got immersed in the project.” For others, the challenge was less about hours and more about uncertainty. “Getting started was the most difficult part,” Aaron noted. “I had never done a competition like this before.” Over time, the team learned not only how to work together, but how to think about data in more nuanced ways.

“This experience taught me that data at face value doesn’t tell the full story,” Zach said. Max echoed that idea from a different angle: “There are multiple ways of interpreting the same data, and there is more subjectivity than is traditionally expected when working with math.” Seeing how the other top five teams approached the same problem during final presentations reinforced that lesson, highlighting that strong analysis can take many forms.

From Nerves to Confidence

By the time the finalists presented, the team had spent weeks refining both their model and their story. Still, stepping in front of judges brought a new kind of pressure. “We felt pretty nervous going in,” Zach admitted. That nervous energy carried through the day. “Focusing during class was hard with the mixture of excitement and nerves,” he said, recalling how he spent every free moment reviewing slides before the presentation.

When the time came, though, everything clicked, and the team put together a final project deserving of the top prize. “In all honesty, that was the best run-through we’d had yet,” Zach said. After presenting, the team felt a sense of calm. They had pushed through setbacks, validated their work, and delivered their strongest performance. “We knew we’d done all we could do and controlled what we could control,” Zach said.

When their team was announced as the winner, the moment carried the weight of everything that had come before it. “It felt really great to win, as we put a lot of time into the competition,” he said.

Looking Ahead

For the Pingry team, the competition was more than a single achievement, it was a glimpse into future possibilities. Several members plan to pursue studies in data science, computer science, applied mathematics, or business, with interests ranging from sports analytics to broader technical and analytical careers. For Zach, the ambition is clear: “My dream is to become a GM one day.”