Big Swing: Robot Sportswriter Outperforms Human When University of Virginia baseball player Will Roberts pitched a perfect game this spring, his feat was covered by two kinds of writers — one human and one electronic. Of the two, the sports website Deadspin determined that the computer better captured the essence of the game.

Big Swing: Robot Sportswriter Outperforms Human

  • Download
  • <iframe src="" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript


A while back, we brought you the story of a breakthrough technology: the robot journalist.

It's actually a software program. You feed it data, it eats that data and spits out a news story just like you would see in your local newspaper. In the beginning, it was used exclusively for sports coverage, and a lot of people were skeptical, namely real-life sports journalists.

Ms. EMMA CARMICHAEL (Writer, Deadspin): I always imagined, kind of the robot you imagined in the third grade with the boxy body and the antennae arms standing in front of a keyboard, like with the beep-boop sound effects.

ADAMS: That's Emma Carmichael, a writer for the sports website Deadspin. And this spring, when she and the other human writers at Deadspin came across an especially bad account of a baseball game on a college sports website...

Ms. CARMICHAEL: Our assumption right away was that maybe this had been written by one of our favorite robot bloggers.

ADAMS: Why? Well, University of Virginia player Will Roberts pitched a perfect game, and that was buried away at the bottom of the story.

Ms. CARMICHAEL: And that was shocking. This was the first time this had happened in the NCAA since 2002. And when it happens, you expect to see it in the headline, and you expect to see everyone talking about that aspect of the game.

ADAMS: Well, the writer of that flawed story, it turned out, was a living, breathing human being. But the creators of Narrative Science, the news writing software, were insulted by Deadspin's assumption. They set out to prove that their program could produce a better story.

Mr. KRIS HAMMOND (Chief Technology Officer, Narrative Science): So we actually got hold of the information director of the school, we got the raw material, the numbers around the story...

ADAMS: That is Kris Hammond, one of the humans behind the machines.

Mr. HAMMOND: ...and we fed it to our system, which wrote the story, where the headline and the lead were focused on the fact that it was a no-hitter. Because how could you write a baseball story and not notice that it was a no-hitter? I mean, what kind of writer or machine would you be?

ADAMS: And here's how the story they sent to Deadspin begins.

MIKE PESCA: Tuesday was a great day for W. Roberts, as the junior pitcher threw a perfect game.

ADAMS: That's NPR sports reporter Mike Pesca doing his best impression of a robo-journalist. The piece goes on and tells the whole game story. Deadspin conceded. They published a follow-up saying that in this case, the machine did write a better story. Sportswriter Emma Carmichael again.

Ms. CARMICHAEL: The image of the robots typing wins me over for sure. And on top of that, in some cases, as we've seen with Narrative Science's story, they actually can produce the stronger story.

Mr. HAMMOND: There was quite a lot of high-fiving and smiling around the office.

ADAMS: You bet. Developer Kris Hammond was on this show a little over a year ago. And at that point, he and his team were hard at work on this crazy sounding technology at Northwestern University.

Since then, they've spun off from the university, formed a startup company, named it Narrative Science, and they say they raised about $6 million from investors.

Mr. HAMMOND: That puts us in a position where we can now grow. And in that year, we've gone from writing baseball stories to football stories to basketball stories to stock reports to earnings reports to market reports for real estate, descriptions of metro areas. And we're looking to a world in which we're going to be able to tell the story whenever there's data out there.

ADAMS: When you're going to sleep at night, do you sometimes say to yourself: This could really be big?

Mr. HAMMOND: So I have a bet with somebody that within five years, a machine is going to win a Pulitzer Prize, and it'll either do so by doing a genuine exhaustive analysis of a dataset that no human being could ever go through and find the nuggets of information and give voice to those nuggets.

ADAMS: So the answer is big.

Mr. HAMMOND: The short answer is big.

ADAMS: A computer program winning a Pulitzer Prize? Maybe it could happen. But Emma Carmichael and the other writers at Deadspin aren't worried about losing their jobs to these machines just yet.

Ms. CARMICHAEL: I think we're happier with the blogger-in-the-basement persona than the robot-in-the-distant-laboratory persona for us. So we'll take that.

ADAMS: And to read the full text of that computer-generated game story we talked about, no humans involved, go to our website,

Copyright © 2011 NPR. All rights reserved. Visit our website terms of use and permissions pages at for further information.

NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.