Abstract: A method is described for obtaining and using viewership data to determine relationships between previous viewership of an episodic program and viewership of future episodes of the program. Aggregating the data may comprise determining missing data points and predicting substitute data points based on past viewership data for the individual viewer and a plurality of other viewers. Data may be presented to the user in a number of ways to aid in analysis and planning of likely viewership for a season of a television program. Data may be used to optimize advertising revenues, plan television lineups to maximize a number of likely viewers, or allocate content and resources between broadcast and video-on-demand.