This monograph evolved from the author's peer reviewed, twenty-first century presidential forecast studies. Across 5 elections, the mean and median errors are 0.6 percent. The author's purpose is to encourage forecasters, political scientists, economists, and sophisticated students to be more innovative forecasters. The only prerequisites are an interest in modern American government and knowledge of basic regression analysis. Dr. Walker enhances Fair's voting model by introducing at least one new issue and additional data for each election. For example, the Korean and Vietnam wars were election years. Enhancing Fair's implicit assumptions reduces forecast errors. The analysis confirms that campaigns in only 7 states plus voting histories in 43 states determine most presidential outcomes.