Technical considerations aside, I think a key question here is whether readers expect just-the-facts stats-based reporting of what happened in the game, or a narrative constructed from the game.
If it's the former, then "sportswriting" about any given game is really just another data delivery/presentation format and should be automated (that, or readers should gravitate towards more efficient data formats, like box scores).
If it's the latter, then a writing program will have to fake creating a narrative, which sounds like a much more difficult job to do. Local sportswriters often know the players, making it far easier to construct narratives around the personalities involved in the games.
My guess is that readers want the latter, but perhaps not enough to pay more, so techniques like those employed by Narrative Science will probably become increasingly common.
Personally, I think the best place for this is in business writing, because I think that eliminating "narratives" from reporting market movements would lead to far more accurate articles.
If it's the former, then "sportswriting" about any given game is really just another data delivery/presentation format and should be automated (that, or readers should gravitate towards more efficient data formats, like box scores).
If it's the latter, then a writing program will have to fake creating a narrative, which sounds like a much more difficult job to do. Local sportswriters often know the players, making it far easier to construct narratives around the personalities involved in the games.
My guess is that readers want the latter, but perhaps not enough to pay more, so techniques like those employed by Narrative Science will probably become increasingly common.
Personally, I think the best place for this is in business writing, because I think that eliminating "narratives" from reporting market movements would lead to far more accurate articles.