Dr. Ross Gore

Dr. Ross Gore

I am Research Associate Professor at Old Dominion University My work focuses on data science and predictive analytics. I am is passionate about extracting useful information from new and diverse data sources to help inform decision making.

AggPro: the aggregate projection system

Published in Baseball Research Journal, 2009

Currently there exist many different systems to predict the performance of Major League Baseball (MLB) players in a variety of statistical categories. We propose, AggPro, an aggregate projection system that forms a projection for a MLB player’s performance by weighting and aggregating the player’s projections from these systems. Using automated search methods each projection system is assigned a weight. The determined weight for a system is applied to all the projections from that system. Then, an AggPro projection is formed by summing the different weighted projections for a player across all the projection systems. The AggPro projections are more accurate than the projection systems when evaluated by average error, root mean square error (RMSE) and Pearson correlation coefficient from actual player performance for the 2008 and 2009 MLB seasons.

Recommended citation: Gore, R., Snapp, C., Highley, Timothy. (2009). "AggPro: the aggregate projection system." Baseball Research Journal. 38(2), 20-25.
Download Paper