Skip to content
Link copied to clipboard

Why the Phillies built an analytics think tank with non-baseball brains

Around the corner from the general manager's office on the second floor of Citizens Bank Park, through the conference room where some of the franchise's most important decisions are made, is a place unlike any other at the ballpark.

Around the corner from the general manager's office on the second floor of Citizens Bank Park, through the conference room where some of the franchise's most important decisions are made, is a place unlike any other at the ballpark.

Six adjustable desks are clustered near a flat-screen TV with a camera for long-distance video chats. Wooden slabs double as countertops at the nearby reading nook. Light enters from the windows that overlook the first-base gate. It resembles the modern offices found at ambitious tech companies, and that is because the man in charge of the room came from Google.

He oversees a five-person staff. One jumped from Northrop Grumman. Another spent a decade with Bank of America before joining Silicon Valley start-ups. One wrote for Baseball Prospectus. One played baseball for MIT, and another just graduated from college in Canada. Four interns arrive in the summer.

The Phillies, once dismissive of analytics, will spend a few million on them this season. Analytics drive more decisions inside Citizens Bank Park than before; they are a component that has gained footing with scouting and player development.

Every team searches for the next competitive advantage. So the Phillies, trying to catch the more progressive clubs, have put a bunch of smart people without baseball backgrounds in this room to study the puzzles of the data-fueled modern game.

It's the Phillies' version of a think tank.

"There's a certain set of internal assumptions of how the baseball world works," assistant general manager Ned Rice said. "For us, that was one of the appeals of getting some people outside the game who really come in with no assumptions. They are all baseball fans to some extent, so they have some baseball knowledge. But they come in with no assumptions. We can critically look at everything and test everything we think about."

Both Rice and his boss, general manager Matt Klentak, are in their 30s. They were hired before the 2016 season to effect change in a rebuilding organization. They would have expanded what was a skeleton analytics operation. But the mandate for a larger investment originated with John Middleton, the most influential ownership presence.

The objective, Middleton said, is a "sustainable competitive advantage" that seamlessly works with the rest of the organization - including the big-league coaching staff.

"The best way to achieve that goal," Middleton said, "is to combine people who are thoroughly grounded in baseball and baseball analytics, Matt and Ned, with extraordinarily bright people who can think critically and creatively and have a proven track record in analytical jobs outside baseball."

Don't fall behind

Major League Baseball's Statcast system recorded 53,380,301 metrics last season from 1,435,241 pitches and 328,405 balls in play, according to Yahoo Sports. The raw data from every night are downloaded to all 30 teams the next morning. Many teams, including the Phillies, have purchased data sets from other sources. There is more information available than ever. The challenge in the big-data age is to translate the numbers into something actionable for executives, coaches, and players.

The next great idea will come from shrewdly interpreting the data, or asking a question that no one else has asked that could lead to a proprietary metric, or examining one of baseball's many uncontested axioms.

"Part of it, at a bare minimum, is making sure you're never behind or missing anything," Rice, 34, said. "You're aware of everything in the public space. You're aware of everything you can be aware of that other teams are doing.

"The second aspect is trying to find ways to get ahead a little bit. A general approach that we think is a little different than other teams do is how many people we've tried to bring in from other industries."

Andy Galdi, 31, is the Phillies' director of baseball research and development. At Google, he was a quantitative analyst for YouTube, and he worked on product enhancements. He interned for the New York Mets before that and worked for the NBA as an officiating analyst. The Phillies declined to make him available for an interview for this story.

The R&D department consisted last season of two full-time employees: Galdi and Lewie Pollis, a 24-year-old analyst who interned for the Phillies, Indians, and Reds and gained a following through his published work with Baseball Prospectus. They were assisted by interns, two of whom were hired this season as full-time software engineers.

The Phillies posted more jobs, and the resumés flowed from other industries. Zo Obradovic became a principal software engineer for the Phillies in February. He came from an Oakland-based start-up called VSCO, which developed apps to edit and share photos. He was a senior vice president for enterprise architecture at Bank of America, then drifted to Silicon Valley, where he was an engineer for Wanelo, an online shopping mall.

Then, Alex Nakahara, a Penn graduate, was hired as a senior quantitative analyst. He spent five years at Northrop Grumman, the aerospace and defense company, as a systems engineer. He wrote a whitepaper on attracting young talent to the aviation industry, which was slow to prevent a brain drain of millennial employees.

"One of the main reasons I joined Northrop Grumman was the mission," Nakahara said in a presentation. "Here you get to work on cool things that make a difference, from space telescopes to advanced aircraft."

Now he'll analyze data trends in baseball and help executives predict which players will rise and fall.

'Get really creative'

It could be the Phillies discover nothing but a marginal advantage from their investment in analytics. The exact work that happens inside the department is secretive; the Phillies, like all teams, are loathe to discuss specifics. Having upgraded from meager analytical insight to a significant amount, the Phillies at least hope to improve their decision-making skills.

Theo Epstein, the Cubs' president of baseball operations and a sabermetric darling, said on a recent CNN podcast that most clubs have "huge" data teams just to stay current. Ten years ago, Epstein could exploit a competitive advantage in the amateur draft with advanced metrics and projections.

"But these days, it's so hard to find any competitive advantage based exclusively on statistics that you have to get really creative," Epstein said. "Everyone has the same information. Everyone is hiring kids out of MIT and Stanford and Ivy League schools with advanced math degrees to dice up the numbers."

Rice used catcher framing as a cautionary example. The on-field baseball people, years ago, extolled the value of an adept receiver. The sabermetricians could not quantify it, so they said it did not matter. Then, when a measurement emerged, the analytics community began to obsess over framing.

"You can get caught up in that groupthink of conventional wisdom," Rice said. "We're trying to avoid that. We're trying to bring in people who think differently."

Epstein hinted that the Cubs have explored medical advantages - either in wearable devices (now approved by MLB), or neuroscience, or predictive metrics for injuries. That is an area the Phillies could mine. Middleton believes the elite analytics departments of the future will rely upon unconventional approaches to common baseball problems.

"We've always been drawn to the idea of hiring people with non-traditional backgrounds," Klentak said, "and allowing outside influences to impact our baseball operation."

Klentak has stressed a sensible presentation of the data, which is all but useless if the impressionable coaches and players cannot grasp it. Those in player development and scouting have begun to breach the think tank's walls. The ideas that happen inside that room, Rice said, must be accessible.

"It's easy for people on either side," Rice said, "to be intimidated by areas they don't know a lot about."

Even the nerds can have a little fun. Two months ago, written in red marker on a whiteboard inside the sleek room, was a statistic familiar to the most casual fan.

Magic number