Newspapers are in deep trouble. That has been clear for some time. It looks like things just got worse for sports writers. The Intelligent Information Laboratory at Northwestern University has developed a program called Stats Monkey that takes the box score from a game and some play-by-play data as inputs and produces a newspaper story about the game, complete with headline. The program uses statistical modeling to figure out what the important events were in the game and to pick out key plays along with a “library of narrative arcs that describe the main dynamics of baseball games.” Here’s a link to a New York Times blog post with a sample of the output from the program based on a recent MLB playoff game.
This is a classic example of substituting capital for labor in production. Just last month, the Washington Times announced a 40% cut in staff, including the entire sports desk. I wonder if they are using Stats Monkey to generate sports page content? I interact with a lot of print reporters, primarily in interviews about my research on the economic impact of professional sports, and have found that the ones working the sports desk generally have the most trouble understanding my research (unlike business desk reporters who seem to quickly grasp the importance of substitution in local entertainment spending for explaining the results in the literature).
The Stats Monkey web site doesn’t appear to list a price for the program, but it has to be a fraction of the cost of a staff of sports writers. I doubt that this sort of capital-labor substitution can save the print newspaper industry, since much of the industry’s problems are on the revenue side of the business.
