A board member at APBRmetrics named "back2newbelf" recently did some research on the viability of our SRS method as a predictor of future games. The result? Over a massive 27,779-game sample, SRS outperforms regular point differential (albeit slightly) in a variety of ways.

This entry was posted on Sunday, January 25th, 2009 at 6:00 am and is filed under Layups. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.