A team of four Butler students reached the final four teams in the undergraduate section of the NCAA Men’s Final Four Analytics Challenge on April 6. Gathered at the Indiana Convention Center, students from 23 Indiana colleges and universities presented data and analytic-based research on NCAA men’s basketball.
The tournament challenges student teams to analyze and create data models around real-life NCAA challenges. This year, teams were given data sets to predict the 68 teams that make the March Madness Tournament, define what creates a championship caliber team and apply it to NCAA business models. Winners received a $2,500 check from the NCAA.
Butler’s team was made up of Tucker Klancnik, a senior finance and business technology and analytics double major, senior computer science major Aman Kikani, Max Feltner, a senior computer science and data science double major and Lucas Krol, a senior business technology and analytics major.
Schools from across Indiana competed in both an undergraduate and graduate section of the tournament. Teams are made up of four to five students, with over 800 total students participating. They are first judged based on their predictive model at each campus, producing one champion of each school. This champion then moved on to face the winners of the other schools, eventually making it down to the final four, where teams were tasked to create a pitch convincing the NCAA to use their model.
This challenge started in 2021, originally partnered with the Crossroads Classic basketball tournament, a series of games that hosted Indiana University Bloomington, Purdue University, University of Notre Dame and Butler. The event has expanded with its partnership with the NCAA coming on as the main sponsor, meaning that the Analytics Challenge now centers on March Madness.
The first step of the challenge this year was creating predictive models through machine learning to predict both the teams competing in March Madness and their potential seeds.
“It was really impressive how close our model was able to get,” Krol said. “67 out of 68 teams [we predicted] before we even knew [the official seeds].”
After making it past the first round, students submitted a technical video explaining their work, with the third round being the presentation. Butler’s team competed in the undergraduate research section against a team from Rose-Hulman Institute of Technology, DePauw University and Notre Dame.
They were the only team to use individual player stats as a part of their research, compared to other schools that only looked at team data points.
“I feel like we knew the most about basketball compared to all the other teams,” Feltner said. “Because we knew [so much] about the sport, [it helped us] build out our seeding, [combined with] pulling external data.”
Professor of business and analytics technology at Butler and Chair of the Final Four Analytics Challenge, Jason Davidson, helped coach Butler’s team. Davidson coordinates with the NCAA, curates the data sets used in the challenge and helps organize all 23 schools competing. He emphasized how important this is for the students participating, comparing the challenge to major U.S. sporting championships.
“Going into the finals [teams] only have 48 hours to solve the last part of the case,” Davidson said. “Our team at Butler probably didn’t sleep. This is our Super Bowl.”
After expanding to the entire state of Indiana, there are potential plans and interest in making it a national competition, with March Madness taking place in Detroit next year. Klancnick, Kikani, Feltner and Krol all encourage students to join the challenge in the future.
“Don’t be afraid to branch out,” Kikani said. “[Even though] I’m a computer science student, doing this sort of thing is a great experience.”
Klancnik reinforced Kikani’s encouragement and explained that it isn’t just about winning a cash prize.
“Don’t be afraid to do it,” Klancnik said. “Be comfortable. Don’t stress about winning. Just have fun with it.”