Follow BallHype on Twitter
www.hardballtimes.com — Monday, October 13, 2008 The problem with measuring forecasting accuracy Posted by Victor Wang at 1:02am David Gassko recently published an article on how various projection systems performed. I am not a big fan of these types of articles. This is not an attack on David or other ...
vote down
The Hardball Times
found this 10/13/2008
The Hardball Times
found this 10/13/2008
2 Comments
-
Derek Carty Great article, Victor. There was actually a similar discussion going on at The Book Blog.
I think it would be very intersting to see how these systems do on players who were projected to be below a certain level (replacement level, maybe?). How many winded up receiving little playing time and how many went on to get a lot of it because they outperformed the projection? I'm not sure of the best scientific way to go about this, there could be some factors you really might not be able to consider (managerial tendecies, etc.) and you could easily end up making some arbitrary benchmarks, but if done correctly, this type of thing would be really cool.
-
Victor Wang Thanks Derek. You're right that a better way of measuring accuracy would to use a method that you describe. However, actually doing something like that would be very difficult. I too would like to see someone try something like that.
-
Related Content
