Jonathan Judge, Schiff Hardin (courtesy photo)
Jonathan Judge never formally studied statistics. He was a piano performance major in college.
And yet, today, the Schiff Hardin litigation partner is lauded as a leading innovator in the stat-head world of baseball analytics. Judge created what many believe is the most advanced, publicly available statistic to measure the true value of pitchers.
Known as Deserved Run Average, the formula for the stat is proprietary and runs 2,000 lines of code long. Judge is in charge of designing statistics for Baseball Prospectus, a site once run by leading statistician Nate Silver and the Internet home for those devoted to sabermetrics, a data-driven movement that has migrated from baseball to other fields, including Big Law. Judge privately consults with Major League Baseball teams who are looking for new ways to value talent.
The Chicago resident does all this in his “off-hours” from life as a partner at Schiff Hardin, a firm he joined in 2003 and where he represents companies facing government investigations into consumer safety issues. For the past two years, Judge said he has spent around an hour a day studying statistics and applying his learning to better understand baseball, as well as fines the government wants his clients to pay.
But Judge’s largest and perhaps most important use of statistics is just getting started: He wants to help the automotive industry describe to regulators, laypeople and juries how the stats-based software behind driverless cars makes decisions. That will be crucial, he said, when cars get into accidents for reasons other than human error.
“Where other people might grab an hour to watch TV or something, I would probably use that time to think about statistical things,” Judge said. “That’s my preferred hobby. I just find it very soothing and interesting and I just find it relaxing, as odd as that sounds.”
Judge’s legal work and his boyhood interest in baseball began to collide in the late 2000s when MLB’s Milwaukee Brewers had a string of winning seasons, culminating in a 2011 trip to the National League Championship Series. A Wisconsin native, Judge had largely given up on the Brewers during their losing years. But their winning renewed his interest in the sport around the same time he began to see a potential use of statistics in his legal practice.
The Schiff Hardin litigator wanted a better approach to arguing over the fines his clients were paying to the U.S. Consumer Product Safety Commission. So Judge figured he would apply some statistical modeling to the history of the fines and use that to ballpark what any new fine should be. He said he could determine a “fair fine” in light of previous penalties with a variation of around $100,000. That would be his starting point with judges and regulators.
“That’s a much better negotiating position than, ‘Please don’t hurt me,’” Judge said.
Regulators have been reluctant to agree with Judge’s approach, he said, for fear that such a rigorous reading of their past actions might limit their discretion in future cases. Such resistance is nothing new for stats pioneers. They’ve been met with wariness almost anywhere they go, often times because the measurements they use can be more difficult to understand than what they’re replacing.
Take Hall of Fame relief pitcher Rich “Goose” Gossage, for instance, who earlier this year blasted “nerds” in baseball.
“These guys played Rotisserie baseball at Harvard or wherever the fuck they went and they thought they figured the fucking game out,” Gossage said. “They don’t know shit.”
The coming battle for hearts and minds over driverless cars is one that Judge wants to lead. And like Silver switching from baseball to politics, Judge is keen on expanding his horizons.
The few accidents that have occurred using driverless technology have made national headlines. The fear among Judge and others is that the public will have a strong reaction against accidents caused by machines rather than humans, despite widespread predictions that driverless technology will lead to less accidents.
As a lawyer with a deep understanding of the data analysis that driverless cars base decisions on, Judge said he can talk directly with engineers and software programmers about why those decisions are being made in the case of an accident. He can then explain it better to regulators, judges and juries in the coming world where he expects car accidents will be treated like a product liability suit.
“There is no one right decision for every situation” a driverless car faces, Judge said. “You can only choose the one that’s the most right the most often. And so you have to be able to explain that. And how can you do that if you can’t even explain how the software is making its decision in the first place?”
Judge has represented a small number of auto companies in this area, but he said the industry largely is still hung up on bigger legal questions around the technology, such as how insurance will work. But Judge expects demand for his advice will pick up quickly.
And so he’ll keep studying statistics while you watch TV.
“I probably should sleep more, that’s for sure,” Judge said.