The National Football League (NFL) concussion scandal led to a $1 billion settlement for former players, a new set of rules for treating head injuries, and even serious concerns about the future existence of the sport.

Much of the credit for bringing this issue into the spotlight has gone to the data analysis of Alan Schwarz, a former New York Times journalist whose series of stories exposing the brain injuries suffered by American football players, and the NFL's attempts to cover them up, earned him a Pulitzer Prize nomination in 2010.

© iStock/fredrocko
© iStock/fredrocko

"The NFL would say that there are a lot of retired players who are having no problems whatsoever. Now, that's not a lie, but it's totally illogical. There are also a lot of people who smoke who don't get cancer," Schwarz told Techworld during the Qlik Qonnections conference in Dallas earlier this month, where he had just spoken about his role in the Data Literacy Project, an initiative that aims to make society more fluent with data.

"History will show I was the only journalist to really call them on that garbage and to look at the conditional probabilities of these guys getting brain disease."

From maths to football

Schwarz calls himself "a proud, card-carrying math geek," whose early ambition to become a teacher led him to study maths at the University of Pennsylvania. When he found out that he would also need a master's degree to teach in state schools, he decided to make a career pivot.

"I went into journalism, which I do believe has a lot more to do with math than most people do," he says. "I think nonfiction writing is actually quite mathematical. It resembles geometric proof, a lot more than most people think."

This belief made baseball reporting a natural place for Schwarz to begin his career. The sport has been infatuated with statistical analysis since "teams were arguing about on-base percentage during the Civil War," says Schwarz. He had spent almost 15 years covering baseball before he first learnt of the connection between concussions and American football through an introduction to Christopher Nowinski, a former college player.

Read next: How data analytics is transforming sports on and off the pitch

After graduating from Harvard, Nowinski established a successful career as a wrestler in the WWE, but it was cut short in 2003 by a series of concussions that left him with symptoms he still feels today. Nowinski had written a book on the subject, and approached Schwarz for advice on getting it published.

A year later, Nowinski called Schwarz with the news that would launch his celebrated investigation. Former NFL player Andre Waters, who had committed suicide, had been diagnosed with Chronic Traumatic Encephalopathy (CTE), a rare and serious form of brain disease normally found in boxers, and more commonly known as punch-drunk syndrome.

Schwartz took the story to the New York Times, where it was published on the newspaper's front page in 2007. By the end of that year, Schwartz had learnt of another three diseased players between the ages of 36 and 56 that had all been diagnosed with that same rare condition. He sought the expertise of scientists who could verify his suspicions. 

"I'm not a professional statistician by any means. But I'm conversant," he says. "I speak the language, and it was very rare for academics or scientists to be called up by a journalist who spoke that language. I think that helped put them at ease that I understood what they were saying and I was going to be able to translate it for the general audience, because so many folks couldn't and did a poor job, so they became very gun shy about speaking to the media."

Verifying the data

Schwarz recognised that there was something statistically unusual about those four players. But the NFL said such a small number could be a coincidence rather than proof.

The denials relied on comparing the results to other players who did not suffer from the condition. But how did they compare to the general population?

"What the NFL and their backers would say was, 'it's just four examples. How can you make any scientific statements based on four examples?' And most people ate that up. How can you say something about 13,000 retired players, if there's only four of these guys? Well, the thing is, the word 'just' was not the right word. It was 'all'.

"Four guys had been examined for the condition - a rare condition to happen randomly. All four had it. Not just four. I was obsessed with those two words. 'Just' was what they said. 'All', is what I said. And what all of it really boiled down to was my proving that that's not going to happen randomly. It's not reasonable to think that these guys got brain damaged in any way other than playing football.

"I don't know how many guys have it. I know it's not a random sample. I know it doesn't mean 100 percent or anything, but there is a higher rate of brain damage among football players than there is in the national population. I'm sorry, don't kill the messenger. It's just true. And the longer you wait to admit it, the more trouble you're going to be in."

The cost to the NFL suggests he's correct. Years later, the league reached a concussions-related settlement with more than 20,000 retired players that was expected to cost up to $1 billion.

It wasn't until 2016 that a senior NFL official publicly acknowledged the connection between the sport and CTE, when Jeff Miller, the league’s vice president of health and safety policy, was asked by the US House Committee on Energy and Commerce if there was a link between football and degenerative brain disorder.

"The answer to that question is certainly, yes," he said.

The next year, researchers at Boston University announced that they had found evidence of the disease in 99 percent of the brains of deceased NFL players donated to scientific research.

Numbers and words

In 2010, Schwarz was a finalist for the Pulitzer Prize for Public Service. Many journalists were surprised he didn’t win, including Malcolm Gladwell, who suggested that the decision was a symbol of American society’s resistance to his message.

He has since turned his focus to educating the media and public about statistical analysis, as a member of the Data Literacy Project's advisory board and the founder of DataPhi Communications, a consulting agency that helps companies convert numbers into compelling stories.

Schwarz is critical of the media’s regular failures to interpret statistics, but acknowledges that those scientists who understand the data aren’t always the best at communicating their findings. His own mathematical training made him an ideal translator of what he calls "the double helix of numbers and words," which he needed to unravel to work out what the data meant. 

Read next: Why we need a right to reasonable inference in the age of big data

"It would be nice if more media organisations and/or editors would allow the calling out of statistical BS, because a lot of editors don't quite get it, either. Some of my editors didn't. They were wonderful, incredibly talented people who could do things I never could, but this is one thing I could do," he says.

He believes that there is an essential need for better data literacy for all members of the public, which may require fundamental changes to the education system. One that he proposes is replacing geometry with statistics in the curriculum.

"If you're a citizen, you need statistics more than you need geometry. I'm sorry, you just do," he says. "If you don't know that the opposite angles of a polygon add up to 360 degrees, I think you'll be okay. I think if you don't understand how compound interest works, you're going to go bankrupt. Which is more important?”