But instead of bulking up their muscles, they were breaking down data.
"They were looking at the springs and the coils and the weight, and I was like, 'What are you guys doing?'" Kelly recalled. "They said, 'We're looking at the coils and trying to figure out the resistance.' I said, 'You're supposed to be lifting the weights, not analyzing how it operates.'
"That made me think: 'Wow. OK, this is what the students at Illinois Tech think about.'"
Brains and basketball since have blended splendidly at IIT, a Division III college known for more for its alumni who invented the cellphone and determined the reason for the ozone hole than for its athletics.
Kelly's lightbulb moment led him to professor Shlomo Argamon and students Denis Bajic and Larry Layne in Argamon's master of data science program who dedicated their summer practicum to basketball analytics.
The result? The Scarlet Hawks have a 17-5 record.
It doesn't take a mathematician to calculate that's a major improvement over last season's 4-21 mark or the 2-23 record they posted in 2014-15, Kelly's first season. The season before that, IIT went 0-25.
Of course, talent and player development are more significant. But Kelly said the number-crunching has made a difference.
"I would say (a difference of) 20 percent," said Kelly, whose Hawks became a Division III program three seasons ago. "It really helps with substitution patterns: Who should be on the floor with whom and which guys play better together. ... I think it's an advantage. Very few schools have a data science and analytics department, especially that offers a graduate degree."
Kelly originally asked the students for information on player efficiency rating — a statistic that measures a player's overall effectiveness on offense and defense — and wins above replacement.
But as Bajic and Layne researched data, they supplied additional information such as players' adjusted plus-minus ratings; the best five-man lineups for defense, 3-point shooting, rebounding and late-game situations; game simulations; and the effectiveness of pairs of players on the court at the same time.
"They had to be creative to figure out how to calculate these standard statistics," said Argamon, who earned his Ph.D. in computer science from Yale. "Then we started asking questions: How can we improve these results?
"One thing we'd look to do is say, here's the best squad for these (specific) purposes. Then I can say, let's put together a good squad not just of good players but players who play well in this particular way."
The May-to-August project was challenging. Because statistics are not readily available for all Division III programs, the students went through play-by-play of games to gather data. Sometimes they would watch video to verify substitutions and other statistical information.
"We had to take a file with all the plays and pull out all the times someone came in and out and when they scored, among other things," Layne said. "We essentially had to write a program that fits itself to a format of the play-by-play data. It's tedious and takes a long time."
Advanced analytics have become a part of the player-evaluation process in sports, popularized through "Moneyball" in Major League Baseball. NBA teams have incorporated more complex data. Division I college programs also have become more reliant on number-crunching, with large staffs of managers and assistants who compile statistics and enter them into programs.
But at the Division III level, few programs have such resources. Besides Kelly, the Scarlet Hawks have only two part-time assistants and one graduate assistant.
"We don't have some of the same facilities," Kelly said. "We don't have some of the same resources. With all those things working against us, I wanted to find something to give us an advantage."
Argamon plans to have students work with the Scarlet Hawks again this summer for the practicum and help garner information for next season. The data should be helpful, considering the low turnover with only one main contributor graduating.
The Scarlet Hawks' success has been satisfying to the academics and athletes.
"When you've quadrupled your win total, you're doing something right," said Bajic, who hopes to pursue a career in sports analytics. "It makes me happy to know the work we put in is paying off. The team is doing spectacular. The school is giving us an education, and we're giving something back. The hope is we can get funding for something like this and have it year-round, not just for basketball but other programs."
Kelly's players are majoring in engineering, business, math, chemistry and computer science. Their analytical thinking helped them grasp and appreciate the implementation of new statistics.
"We have a lot of bright people at our school and on our team," Kelly said. "They're more used to data and analytics and numbers. They always want to figure things out."
As the team rides a seven-game winning streak, the players apparently are solving things on the court too.
©2017 the Chicago Tribune Distributed by Tribune Content Agency, LLC.
