This article was published on February 12, 2013

Automated Insights uses MLB’s big data to generate slick game feeds, replays and commentary


Automated Insights uses MLB’s big data to generate slick game feeds, replays and commentary

Big Data company Automated Insights is launching a slick demo app today at the On Deck sports and tech conference based on its ‘artificial intelligence platform’ for real-time data. It’s called Real-time Insights for MLB and it presents games from the 2012 season in an interactive format that allows you to watch them happen in “real-time” (actually slightly faster).

The slick bit here is that it presents the events of the game as they happen, but also generates automatic natural language ‘insights’ along with it that mirrors the kinds of facts you might hear an announcer give you during a game. For each play, real-time notes are created and assigned scores that rate their ‘importance’. They’re then presented in order of ‘depth’ below each play.

So, you might see “Jose Altuve struck out swinging. None on with one out and Brian Bixler due up.” as the play, and then below it you’re presented with the following insights, in order of importance:

  • In his career, Madison Bumgarner has struck out 350 opposing hitters.
  • Jose Altuve has struck out at least once in 26 of his 60 games this season.
  • That’s the first strikeout of the game for Madison Bumgarner.

These are the things that an announcer might choose to share with a person watching the game, as it gives them more context as to the importance of that particular play. It’s all pretty darn sweet to be honest.

Screen Shot 2013-02-12 at 11.52.16 AM

The Real-time Insights for MLB product will end up being a feed that media companies can buy and tuck into their web presences or apps. The demo app described above, and seen here, is just one of the ways that Automated Insights can leverage and parse the data, but it’s an impressive one.

The app will be used for live, real-time feeds of insights during every MLB game this year. It will use play-by-play data and a database of stats going back 120 years to generate the realtime predictions and facts that you see attached to each. It’s not hard to imagine the possibilities that this could bring to sports apps, I would love to see this integrated into the already great MLB At Bat, for instance.

The MLB app isn’t the end of it though. Robbie Allen, Automated Insights CEO, said at the conference that “in five to 10 years, we are going to look back and think it silly we ever had humans on data analysis.”

Other major sports, finance, traffic and weather applications of Automated Insights’ platform are on the way as well. We’re just seeing the beginning of the way that real-time data is combined with big stores of historical data and parsed in natural language.

Image Credit: TOSHIFUMI KITAMURA/Getty Images

Get the TNW newsletter

Get the most important tech news in your inbox each week.

Published
Back to top