John Savage stood in the third base dugout at TD Ameritrade Park looking at his clipboard. Baseball moves slow like the sun, but decisions happen fast. And the UCLA coach, locked in a close CWS opener, had to make a call.
One of LSU's best hitters was at the plate. All of Savage's baseball training told him to pitch around the batter. Go after the next guy. But the numbers told him something else — the Tiger at the plate hadn't shown power against righties.
So Savage let his ace attack. And Adam Plutko got the out.
UCLA rolled to five straight wins in Omaha because its pitchers hit spots, its defense mastered fundamentals and its offense did just enough. But the Bruins also had a slight edge.
Behind Savage's clipboard was an unusually large database of advanced statistics. And behind that database is an Omaha native who got hooked on “Moneyball” in — of all places — Washington, where he worked for a prominent U.S. senator. Now Kent Bonham is helping lead college baseball into a new era.
The website Bonham and colleague Jeff Sackmann launched, CollegeSplits.com, is one of the latest ripples in the sabermetric wave sweeping America's pastime. UCLA is one of 15 college baseball programs who pay for complete access.
“I really think it's been a big part of our success,” Savage said Friday.
UCLA doesn't have time to thoroughly scout every opponent, especially in the postseason, when it faces teams across the country. But Savage can check College Splits and see exactly what every LSU batter does against left- or right-handed pitching. How many ground balls a North Carolina pitcher induces. What percentage of a Mississippi State hitter's at-bats are fielded by the right fielder. And much more.
If he fails to teach and prepare his players, Savage said, the numbers won't save him. But having just a little more information helps him formulate strategy and gives his players confidence that he's basing decisions on more than a hunch.
“It's a game of leverage, a game of odds,” Savage said. “The more I know, the better.”
* * *
Since Billy Beane's “Moneyball” went mainstream 10 years ago, professional baseball analysis and vocabulary have changed dramatically. Experts are now just as likely to cite OPS (on-base percentage plus slugging percentage) as batting average. More likely to reference WAR (wins above replacement player) than all-star appearances.
The idea behind the new numbers is to evaluate what's really going on. Remove bias and intangibles. Compare individual contributions objectively.
If a team wins 90 games, where do those wins come from? The shutdown closer with 30 saves? The second baseman with only three errors? The first baseman with 40 homers? Through (sometimes) advanced formulas, sabermetricians can place values on strikeouts, doubles, stolen bases, most anything.
Their research has established new guidelines in player analysis. For instance:
» Don't buy into stats like RBIs and ERA; they depend too much on factors beyond their control.
» Don't give away outs on the basepaths.
» Don't give the opponent a free base — no intentional walks!
» And perhaps most of all, a sacrifice bunt is a bad idea 99 percent of the time.
Baseball insiders, especially those who inhabit the dugouts, have traditionally rolled their eyes at the stat geeks. They say the game is too complicated to reduce to formula. They mock the notion that a stranger with a calculator knows more than a coach.
“How do I manage?” said Indiana coach Tracy Smith. “I manage with my gut. I'm a simple human being. And I've been blasted on Twitter. I've got guys questioning, 'Coach, why are you bunting first and second with nobody out?' And they cite all these sources that say that's the absolute stupidest thing in the world. I'm just a big believer in I think you've got to know your personnel.”
Winning arguments isn't why Bonham and Sackmann launched College Splits. Like any die-hard sabermetricians, they don't mind debunking myths and slaying sacred cows, but scouts vs. stats, Bonham said, is a “false dichotomy.” Analyzing college players should always include both. Only one of those components, however, is trending up.
Recently, a draft analyst came to College Splits inquiring about a top prospect from a small school. The player had a reputation, based on scouting assessments, for struggling against quality pitchers. Bonham checked the data.
“Turns out, he performed just as well on Friday nights against No. 1 starters as he did any other night.”
* * *
Bonham remembers parking behind the left-field bleachers at Rosenblatt and hiking up, up, up the hill to the College World Series.
He grew up just north of 108th and Center. Graduated from Westside in 1990, then studied government at Colby College in Maine. He came home in '95 and started looking for a job. In the interim, he volunteered for an investment banker's upstart Senate campaign.
Nobody knew Chuck Hagel. Nobody thought he would win. But Bonham worked a few hours a day stuffing envelopes and making coffee — “I had nothing else to do.”
One night, Hagel's campaign manager called Bonham's house looking for him. Hagel's driver wasn't available the next morning. Could Kent pick up Hagel early and chauffeur him to events?
That was his job for almost a year. When Hagel upset Ben Nelson in '96, the 24-year-old Bonham followed the senator to Washington, where he rose to legislative director and, ultimately, deputy chief of staff.
The hours were long, the responsibilities serious, the pay low. Bonham found stress relief in baseball. A Red Sox fan since college, he got hooked on the analytical side of the game. While his colleagues were keeping tabs on Frank Solich back home, he was reading the godfather of sabermetrics, Bill James.
Is there a better way to measure defense than fielding percentage? Is there a better way to measure pitching than win-loss record? Should Bill Mazeroski and Catfish Hunter really be in the Hall of Fame?
In 2002, John Henry bought the Red Sox and tried to hire Oakland GM Billy Beane. It was big news in baseball.
Why, Bonham wondered, were Beane's ideas so revolutionary? Bonham dug deeper, devouring the newest studies at online sites like Baseball Prospectus and the Hardball Times.
How you compare offensive numbers across different eras? Is there really such thing as a “clutch hitter?” And who really is a team's most valuable player?
Decades-old pillars of baseball theory were crumbling under a pile of new research. If there's a tough hitter at the plate and first base is open, conventional wisdom said walk him. Research said otherwise. Conventional wisdom said put your fastest guy at the top of the lineup, even if he doesn't get on base much. Research said otherwise.
In 2001, Bonham left politics and took a job at a D.C. consulting firm. He started a family, then started chasing an MBA from Georgetown.
When his baby girl finally fell asleep at 10 or 11 p.m., he grabbed his laptop, opened Microsoft Excel and toggled between his finance homework and college box scores. His initial motive was curiosity.
How much does the North Carolina ace's fastball really sink? He went to the UNC website. Found the play-by-play sheet for every start. They didn't show where a batted ball landed, but they did show who fielded it and whether it was a ground ball or fly ball. With that information, Bonham could crunch numbers.
Trouble was, it took too much time.
Enter Jeff Sackmann, who was making a splash in the sabermetric world with a site called Minor League Splits. He had created a software program that automatically compiled and broke down hundreds of prospects by their splits — righty/lefty, ground ball/fly ball, home/away, runners in scoring position/nobody on base, etc.
Bonham's review: “Holy crap, this is incredible.”
He sent an email to Sackmann, whom he calls “insanely smart.” They struck up a conversation and, in 2006, met for lunch in downtown D.C. That's where Bonham pitched an even bolder idea.
Let's take the same blueprint and apply it to college baseball.
* * *
For years, evaluating college players had been an even greater mystery than minor-leaguers. How can you possibly compare one prospect who plays for Miami versus one who plays for Air Force? The season is short, the schedules and ballparks aren't close.
A prospect might go 4 for 5 against a Division II school on Tuesday, then go 0 for 5 on Friday night in the Big 12. That's not a true .400 average, is it? Statistics were almost useless. As a result, evaluations consisted primarily of a scout's report, based off a few random games.
Bonham and Sackmann spent several months studying, planning and constructing their site. It was a personal challenge more than a business opportunity — they were just trying to break new ground in the sabermetric community.
On Feb. 9, 2007, Sackmann unveiled CollegeSplits.com
“Literally less than 5 minutes after we published something,” Bonham said, “major league teams came knocking.” They included his sabermetric inspiration — Billy Beane.
“The data Jeff and Kent provide has the potential to revolutionize the way teams analyze players for the draft,” Beane says on the College Splits website.
A pro scout could study swing mechanics and determine if a prospect effectively hit the ball to the opposite field. But it was no longer a subjective piece of analysis.
“We have the data,” Bonham said. “We can tell you if that's true or not.”
College Splits had immediately found a niche — more than half of major league teams are currently clients. But a breakthrough came a few years later, when the first college coach reached out. If front-office personnel could use the splits to evaluate prospects, why couldn't John Savage use the numbers for scouting opponents?
He still uses video — lots of it. But the numbers are invaluable, Savage said. One of the toughest things in college baseball is gathering information on opponents. Too often you go into games blindfolded.
“We struggle when we really don't know who we're playing,” Savage said. “I'm telling you, that is a terrible feeling for me. A terrible feeling. We can get beat by anybody we don't know.”
But when Savage sees an opponent hitting .380 against lefties and .220 against righties, suddenly he has an edge.
There are legitimate reasons to be skeptical of splits. No. 1 is sample size. A major-league season consists of 162 games. Some three-year college players don't play that many games in a career. Does a coach want to make a decision based on a handful of at-bats? Beats the alternative, Sackmann said.
“If somebody could tell you that your player hits better when he's batting leadoff than when he's batting seventh, you might have good reason to go against that,” he said. “But it's better that you know.”
Sackmann does the heavy lifting on College Splits, overseeing a database with more than 125,000 players, crunching the numbers in manageable bites. The site is updated daily.
Bonham, the Omaha native, maintains a full-time career in financial regulation — “This is kind of the best of both worlds. I have no desire to work in the game.”
He does make an annual trip to the baseball winter meetings to meet with clients — and court prospective ones. He also keeps an eye on his hometown every June.
Tuesday night, seven years after he pitched his big idea, Bonham watched the ninth inning from his Washington, D.C., living room. The house was quiet, the kids were in bed, the work was done. It was pushing midnight when UCLA recorded the final out. As fireworks burst outside the stadium, the Bruins dogpiled behind the pitcher's mound, celebrating their first national championship.
It was Bonham's first, too.