Graphic by Deposit Photo
“North Carolina grabbed the lead in the top of the 10th inning as a wild pitch by Clark Labitan allowed Colin Moran to score the go-ahead run. The Tar Heels held on to defeat Virginia Tech 9-8.”
Jim Murray, it is not, but dig a bit deeper and the significance of this lede comes into clear focus. It wasn’t written by a reporter who covered a game. Instead, it was composed by a computer.
Later this month, the Associated Press will be churning out similar computer-generated ledes and stories on college baseball in a new deal with the NCAA. The pact calls for AP to employ “automation technology” to cover college sports beyond big-time football and basketball, including those at the Division II and III level, that traditionally don’t receive coverage.
Hold your breath sports journalists, because it’s just a start, says Lou Ferrara, vice-president of sports, business and entertainment news for the Associated Press. He says computer-generated game stories eventually will make their way to the bigger sports like Major League Baseball and NFL games.
“That’s our next wave,” Ferrara said.
Later Ferrara added, “I look at sports as a pivot point of change [for automation technology].”
The technology was developed by Automated Insights, a company based out of Durham, N.C. It relies on a Wordsmith platform that can analyze data from a game and turn it into a game-story or a recap.
Robbie Allen, the founder and CEO of Automated Insights, says the program could pinpoint that a double in the eighth inning or a missed free throw down the stretch proved to be key pivot plays in a game. STATS, a Chicago-based company that feeds sports content worldwide, saw the potential for the technology when it purchased Automated Insights earlier in the year.
Previously, Yahoo! Sports has used Automated Insights’ technology for fantasy football reports and previews and recaps of events. The AP also uses the technology for computer-generated business reports.
The NCAA deal, though, will be its largest application as it relates to sports.
“We have many customers who want coverage (of college sports beyond big-time football and basketball),” Ferrara said. “The cost to staff all of those games would be huge. Now we have a low-cost solution to provide a high volume of content. The local markets will benefit from coverage they never got before.”
Ferrara insists the computer-generated game stories are a natural evolution in sports coverage in an ever-changing media landscape.
“The traditional game story is an endangered species,” Ferrara said. “In an age of Twitter and broadcasting [of multiple games every day], the value of those traditional game stories has declined for our customers.”
The notion of computer-generated game stories probably isn’t going to give sports journalists a heightened sense of job security. We’re sorry, but you’re being replaced by a Mac.
Ferrara, though, stresses the new technology isn’t about eliminating jobs. Rather, it is about using reporters’ time in a more efficient manner.
“AP still will be at every NFL, MLB, NBA and a lot of NHL games,” Ferrara said. “That won’t disappear. But we need our reporters on stories you can’t get anywhere else. That’s what people want. Eventually, as the game stories become automated, it will allow them to focus more on that reporting…The reason we all get into this business is not to process data; It’s to report, break news and tell stories.”
Added Allen: “Our biggest value is augmenting what’s already there. It’s not replacing what they already provide.”
But what about those game stories? Even if their emphasis is being reduced, they still serve a purpose. There are nuances in a game that won’t be picked up by a computer.
For instance, Allen acknowledges a computer won’t report an outfielder lost a flyball in the sun, allowing the winning run to score.
“Our stuff is quantitative related,” Allen said. “We’re not able to make a statement about the quality [of a play].”
Ferrara has heard similar concerns from sportswriters. A computer also isn’t going to delve into critical coaching and managerial situations that impact a game.
“I say if that happens, we do what we do now. We go and report it,” Ferrara said. “We’re still going to have reporters asking the questions. That becomes the story.”
A computer can’t write like Grantland Rice and deliver an epic lede that begins “Outlined against a blue-gray, the Four Horseman rode again.” Then again, nobody has written like that in more than 80 years.
The future is here, and Ferrara maintains the computer “only will get smarter” when it comes to automated-game stories. Its use only will become more wide-spread.
“Everyone sees the automation of these game stories in the future,” Ferrara said. “If you’re the NFL, why wouldn’t you use it? It makes sense.”
————————————–
Recommended reading on sports journalism:
Michael Bradley and yours truly weigh in on the Bill Simmons fallout at the National Sports Journalism Center at Indiana.
Welch Suggs, writing for Medium.com, gives his reasons why he is not a fan of Simmons.
Jeff Pearlman did an interview with CBSSports.com baseball writer Jon Heyman.
Ed Sherman writes about sports media at shermanreport.com. Follow him @Sherman_Report.
Comments