from Amazon
Wednesday, October 14, 2009
Setup for Forecast Evaluations
The kids who made the Steamer Projections are now starting to work on evaluating the success of several projections systems for the 2009 season (including their own). They’ve read Nate Silver’s evaluation of his system after 2007 as well as your subsequent back and forth with him (http://www.insidethebook.com/ee/index.php/site/comments/forecast_evaluations/) about how to do it right. [Editor’s note: yes, that’s the best thread that discusses the best way to do the evaluations.]
Anyway, they’re lining everyone up using Lahman IDs and would be happy to include any system that has Lahman ID’s in their analysis. I see that Marcel, conveniently, already does, but I was wondering if other forecasters
might be interested and/or if you might still have their systems with Lahman IDs lying around. If you post this or send it to the other Forecasters in the roto challenge we’d be happy to take suggestions of comparisons people
might want to see.They’ve got the Marcel, ZiPS, Pecota in excel with Lahman ID’s right now and are looking to add Sporting News and ESPN. I’m going to suggest that they also use CHONE and maybe OLIVER so they can compare results to Matt Swartz’s work last year (http://statspeak.net/2009/04/testing-the-projection-systems-strengths-and-weaknesses.html).
Email professor Jared at: jfcross4~gmail~com , replacing the ~ as appropriate.
Recent comments
Older comments
Page 1 of 344 pages 1 2 3 > Last »Complete Archive – By Category
Complete Archive – By Date