If money were no object, there is one element Brad Stevens would add to Butler's basketball program. Not an opulent practice facility. Nor a university jet to transport the Bulldogs to road games.
"I'd probably create a statistics division," the coach said.
Analytics have gained widespread acceptance in many areas of business, and Stevens, a self-described stats geek, has been at the forefront of the movement in college basketball. Stevens, who graduated from DePauw University with an economics degree, uses numbers as a tool to prepare the Bulldogs for games and evaluate their development.
That data analysis is one reason Stevens, 35, led Butler to back-to-back NCAA championship games.
"Almost everyone uses some of this," said Sam Hinkie, an executive vice president of the NBA's Houston Rockets. "He does embrace this in a way that differs from most."
Butler, which opens its fifth season under Stevens on Saturday at Evansville, has drawn comparisons to the Oakland A's of "Moneyball," and Stevens has been likened to A's general manager Billy Beane. Bobby Fong, the former Butler president and a longtime baseball fan, made such an analogy nearly three years ago -- before it was trendy and Brad Pitt starred in the movie.
"What struck me even three years ago, in college basketball, defense was undervalued," said Fong, now the president at Ursinus College (Pa.). "There was an emphasis on defense first in the Butler system."
The A's of Moneyball used unconventional data to locate players who were undervalued and thus lower-priced during a four-year playoff run from 2000-03.
The Butler parallel is not precise because it uses statistics largely to prepare for opponents, not for player transactions. Indeed, Stevens suggested Butler might be the anti-Moneyball in recruiting because it does not rely on high school statistics. The only high school numbers that merit attention, he said, are 3-point and free throw percentage.
The Moneyball analogy is applicable in that the Bulldogs search for undervalued skills or prospects. While some programs recruit athletes as if they're trying to win track and field medals, Butler identifies those who fit its system.
But if there is a way to quantify anything in basketball, Stevens uses the data.
"You're so competitive that you just want to find any little thing that might give your guys an advantage out on the floor," he said. "And it's the way I'm built and driven. I really did enjoy the book 'Moneyball' when it first came out. I read it right away."
kenpom.com
Stevens is guarded about some details of Butler's analytics, not wanting to forewarn future opponents or disadvantage any active players. But there is no secret about where he starts opponent evaluation -- kenpom.com.
Ken Pomeroy, a meteorologist living in Salt Lake City, started a website in 2004 that introduced unconventional data to college basketball. Pomeroy assembles stats such as points per possession, percentage of offensive rebounds and ratio of 3-pointers to total shot attempts. Stevens examines kenpom for a snapshot of tendencies.
"He's obviously made my site more famous," said Pomeroy, who has never spoken to Stevens.
Kenpom is complemented by Synergy Sports Technology, a video scouting system that can isolate player tendencies. Stevens said he wants film to validate what stats are indicating. Ninety-nine percent of the time, he said, they coincide.
"Data can bridge that gap between the scout and the dozens and dozens of games he didn't get to see," the Rockets' Hinkie said.
Statistics are misleading if the wrong ones are studied. Stevens said season stats won't reveal the right information if one player, for instance, has been hot in recent games or if the team has changed the way it plays.
Stevens also charts what he calls "typical scoring." In other words, he might not change defense against a hot-shooting opponent, thinking, "The law of averages may just kick in here."
A scouting report also can't prevent the inevitable. During Stevens' first season on staff, 2000-01, Butler determined what the preferences were for 6-8 power forward Michael Wright of Arizona. But Wright still overpowered the Bulldogs inside and shot 10-of-12 in two games.
"He buried us so deep, it didn't matter what his preference was," Stevens said. "He was just going to lay it over the rim."
Moreover, a detailed scouting report is meaningless if it is too complex to absorb. More than anything, Stevens said, players must be able to carry out a game plan.
"It's not what we know. It's what they know," he said.
Details
One of the reasons Butler players have approached NCAA Tournament games with such confidence is trust in the game plan. Stevens has often told them what would happen in a game . . . and players have watched action unfold exactly as he described.
"There's no one in the country who pays closer attention to the details than him," senior guard Ronald Nored said.
Stevens has motivated the Bulldogs by producing statistics showing they're not as bad as critics might claim. Or the coach has demonstrated how a small statistical improvement -- like two fewer field goals per game by an opponent -- would result in a big difference.
Sophomore forward Khyle Marshall said "too much at once" is hard to process but that "bits and pieces" are not.
"As it just expands and expands, it's something we have in our minds," he said.
Stevens focuses on details because Butler rarely wins via blowout. His favorite example is from the 2008-09 season: If Butler's 13 victories by three or fewer possessions had been defeats, the record would have been 13-19 instead of 26-6.
The Bulldogs sputtered in each of the past two seasons before putting together 25- and 14-game winning streaks. To Stevens, a team is constantly evolving. The Bulldogs won't be as good in November as in February or March, he said.
"I like that. I like getting better," he said.
At Butler, it has all added up.
Butler basketball coach Brad Stevens studies advanced statistics to give his team an edge, similar to the Oakland Athletics and general manager Bill Beane as chronicled in the book and movie "Moneyball." Here are some examples:
-Old Dominion, the Bulldogs' March 17 opponent in the NCAA Tournament, led the nation by rebounding 45 percent of its own missed shots. Stevens called it as "staggering a number" as he had ever seen. Keeping Old Dominion off the boards became an emphasis for Butler, which led 32-29 in rebounding and won 60-58. "Butler cares about defensive rebounding," Houston Rockets executive Sam Hinkie said.
-Old Dominion guard Kent Bazemore drove to the right 75 percent of the time (even though he is left-handed), according to Synergy statistics. "If you say that somebody likes to go right, well, document that he likes to go right," Stevens said.
-In the 2010 NCAA Tournament, Butler players watched film of Syracuse guard Andy Rautins shoot open 3-pointers and vowed that he would not do so against them. Butler calculated that Rautins shot 80 percent of his attempts behind the arc and moved in one direction -- Stevens wouldn't say whether it was left or right -- 70 percent of the time. "That simple fact is hard to forget," Stevens said. After scoring 24 points in his previous game, Rautins had 14 as Butler upset the top-seeded Orange 63-59.
-For a poor offensive rebounding forward, Stevens said, "You're beating a dead horse if you say, 'Go to the glass, go to the glass.' " Instead, the coach said, show that he went to the boards three times out of 50. Either the forward is not in shape or not trying. "One more possession might mean winning the game," Stevens said.
-When players do conditioning drills in the fall, they are timed in four sprints. Stevens cares more about the difference between the first and fourth than in the average "because that shows how in shape you are," he said. Those who need more conditioning get that, and others concentrate on developing skills.
-Butler has 11 years' worth of data on what a typical All-Horizon League guard would score in conditioning tests. "So you try to motivate with those numbers," Stevens said.
-Matthew Graves, the Bulldogs' associate head coach, oversaw three years of workouts in which Matt Howard shot 25 to 50 3-pointers per session. It was all charted. "For two years, in individual workouts, we had the confidence and he had the confidence because we had the data to back that up," Graves said. Howard, who made five 3s in his first three years, sank 53 last season and shot 40 percent from the arc.
