StatSheet, a Durham, N.C., company that serves up sports statistics in monster-size portions, thinks otherwise. The company, with nine employees, is working to endow software with the ability to turn game statistics into articles about college basketball games.
This strikes me as a novelty. The text produced by the software is terrible to read and the analysis itself is hardly worth reading it for. “Over the last four seasons, Washington is 1-0 against Long Beach State,” goes the final sentence of this especially compelling game preview. But it does make for a worthy discussion point.
If computers and the internet brought the democratization of sports news and analysis, then why not also the mechanization? Is this a logical progression? Are robots capable of providing not just information but analysis that people actually want to read? The CEO of StatSheet obviously thinks so.
Gerard Cosloy of Can’t Stop the Bleeding jokes that one day soon we’ll be longing nostalgically for Bleacher Report, with its tremendous brand sporting insight. My thought is that integrity-wise, the robot sites aren’t too far removed from Bleacher Report –a site that, by all appearances, levers cheap and free labor into page views, often at the expense of decent content. But overall, I tend to agree with Cosloy’s facetiousness.
Robots can’t win — at least not yet. The reason is complexity. As the linguistics expert in the Times story notes, robots can’t write nuanced sentences. Bleacher Report writers (in theory) can. And for sports analysis to be even decent, it has to do more than spit statistics and canned catchphrases out in seemingly random order (although this strategy works very well in radio). Decent sports analysis is about ideas and perspectives — this is true whether the analysis is wonky like the work on FanGraphs or Ken Pomeroy’s site, or news and gossipy like much of, say Bleacher Report’s content, or generalist and frequently long-winded like the writing this website.