Please create an account to participate in the Slashdot moderation system

 



Forgot your password?
typodupeerror
×
Robotics

An Algorithm May Soon Cover Your Local Sports Team (vice.com) 53

Sam Edwards, writing for Motherboard: A Spanish startup is promising to revolutionize readers' access to often unreported news. The unreported news in question, however, is not overlooked disasters or under-reported tragedies in far-flung countries, but minor league sporting events. David Llorente, co-founder of Narrativa, said was inspired to develop an AI-powered content generation system after he tried fruitlessly to find coverage of minor league soccer games from other countries in his native Spanish. "There are people interested in these things, in these leagues, in these kind of sports," he told Motherboard. "The idea was to focus on regional sports. I wanted to write about football, but about Japanese football in Spanish, to cover this niche." Sevilla won with a resounding 20 against Athletic in Nervion, where the sum up eight straight wins at home. Gameiro scored the first one for the locals and closed the scoreboard by converting a penalty kick after Kychowiak was fouled. Athletic was unlucky despite controlling ball possession and wasn't able to finish any of the numerous chances that they had. -- Narrativa game summary.
Narrativa is part of the booming automatic content generation industry which uses algorithms to convert data sets into narratives.
Related: How a robot wrote for Engadget.
This discussion has been archived. No new comments can be posted.

An Algorithm May Soon Cover Your Local Sports Team

Comments Filter:
  • Comment removed based on user account deletion
  • Go sports team! Do the sports good and beat the opposing sports team!

  • Once again, sports -- and by extension sports commentary -- is a form of artistic expression (outside of the business of sports, of course). If an algorithm can give me the commentary, then I'm not interested in that commentary at all. It doesn't express a human-art, and therefore it contributes nothing of value to my day.

    • by DogDude ( 805747 )
      What could possibly happen in sports that hasn't happened a million times, already? What could be so interesting as to require a human to explain it? Listening to people "report" about sports is already mind-numbing, because people just say the same stuff over and over. The players "really wanted to win"? The team "gave 110%"? The coach is "disappointed that they lost"? The person kicked/threw the ball really far? I would argue that there isn't any "art" in reporting on sports at all, because there's
      • by Sowelu ( 713889 )

        I'd say there's art in telling (some) stories, even if the listener has heard it before. YMMV which ones are interesting. Helps if you have emotional investment in one of the teams, whether there's any point to it or not.

        • My point is that in those cases, it's the "telling" that's art, not the "hearing". When an algorithm does the telling, there's no art left.

          I really don't care about what a computer has to say about a game. For that reason, I'm not interested in hearing it.

          • by Sowelu ( 713889 )

            I'd argue there's an art in designing the algorithm. At least I'd hope so, since I've got a hobby project involving algorithmic writing. I might be biased.

            • Oh, certainly. But that's for the documentary on how-it's-made, not for the consumer listening to it.

            • I'd argue there's an art in designing the algorithm.

              I doubt if there is any art in it. You just take the statistics of a zillion sports games, and the articles written about them by humans, and shove it into a ML system. Then cycle through them until the system learns to generate reasonable articles from the data.

      • Well, perhaps you forgot that just last week, Facebook's AI picked up a fake news report & ran it as real news. Just days after they fired their human editors.

        So, there's all kinds of hilarity to be expected.

    • Once again, sports -- and by extension sports commentary -- is a form of artistic expression (outside of the business of sports, of course). If an algorithm can give me the commentary, then I'm not interested in that commentary at all. It doesn't express a human-art, and therefore it contributes nothing of value to my day.

      I wouldn't expect to be uplifted by the soaring prose of a minor-league sports report, unless they've singled out an exceptional minor-league game, which is a different goal from this project. For run-of-the-mill sports reporting, doesn't most of the "art" consist of identifying important or unusual details of the game -- which play changed the momentum from one team to another, which player(s) performed better or got more playtime than they usually do, how well or poorly the great player(s) performed? That

    • You could replace many commentators with robots and not miss a thing. Commentating is 80% filling dead air with cliches 19.999% relaying what just happened and 0.001% actual insight.
      • But that's exactly what makes it an art. It's about choosing which 80%, which 19.999%, and which 0.001%. That's the expression.

        This is slashdot. Every week we read another "high school class sends camera into space for under $100, gets photos as good as nasa" article. But that high school class got 100 random photos, once. Nasa gets the photo they wanted, of the object they wanted.

        It's the very selection that's the art.

  • I'm ashamed at how many hours I wasted on the 2012 version of that. It was pretty good at algorithmic descriptions most of the time, but its stockpile of phrases during matches left a lot to be decided. I shouldn't see "He puts the ball in row Z!!" several times a match, especially when someone just barely gets a ball into the stands. Oh well.

    I keep thinking "This year's version will be better", then remembering the life I got back when I stopped playing.

    • by Sowelu ( 713889 )

      ...left a lot to be desired. Sure does take a long time to wake up after a three day weekend.

  • No, thanks. Just give me the scoreboard and stats and I can read the data myself. I don't need a robotic overlord to dumb it down into humanspeak.
  • Pretty sure Narrative Science has been doing this since 2012 [wired.com]. At least for Little League.

    Also, their competitor Automated Insights offered API access to small parties last year [wired.com]. Maybe "local sports" is too big?

    Maybe this is new for Spain?

    Regardless, seems like entry level writing positions are going to be more difficult to come by, at least for humans.

  • The problem that I see is that good commentary creates the narrative of the game. Sports has not actual intrinsic stakes for most fans (short of a few bets here and there), but the commentators and news sources allow for us to be fed a narrative of how much the underdogs have overcome by strength of will to make it this point, etc etc. I question the current AI's ability to do this coherently and not just report who won and what happened. Because in general, that's rather uninteresting.
  • Somebody - or something - does not not the first thing about football. Or soccer, as it is called this side of the pond.
  • So it takes perfectly good statistical data and turns it into wordy, clichéd prose? What will retired althetes do now?
  • by JustNiz ( 692889 ) on Tuesday September 06, 2016 @03:39PM (#52837101)

    Easy for my favourite sport:

    while ( stillRacing )
          {
          printf( "Hamilton is 1st, Rosberg is 2nd, Button is nowhere to be seen\n" );
          sleep( 10 );
          }
    printf( "Hamilton won\n" );

  • If an algorithm can produce the commentary, then a format more orderly than paragraphs of text (e.g., a box score) will convey that same information better.

    Most things said and written about sports are vapid by Sturgeon's Law.

    • Most things said and written about sports are vapid by Sturgeon's Law.

      I think it depends on the context. A lot of sports reporting is high level fluff that can probably be automated. But my chosen code, I follow ex-players and coaches who get semi-privileged information directly from their peers they used to share the locker room with. They also have insight into the nuances a desk journalist or robot just can't cover.
      So there is an evolution there. General results will be covered by robots, but if you want the gritty emotional details you'll still need a human.

  • And it doesn't seem worse than any other method.

    In fact, ESPN seems to be doing this already - in NFL news their regional, conference, and team news spews regularly take on the same flavor, for instance, a topic of 'Top Five Special Teamers' or something similarly predictable and generic will pop up, especially in the off-season when there is, in fact, a lack of 24 hour cycle news.

    Blah. The spew is already robotic. Just dispense with the meat robots.

  • How many people want to read about Japanese football in Spanish? I understand people of the country wanting know about the regional sport in their country but not the regional sport from halfway around the world that isn't even a hub for that particular sport.

Every nonzero finite dimensional inner product space has an orthonormal basis. It makes sense, when you don't think about it.

Working...