This program shows how statistics are used to forecast the outcome of elections, and how faulty polling methodology produces inaccurate results. The program focuses on the 1992 British parliamentary elections, in which four major polling companies predicted the wrong outcome. We learn how data was distorted by poor sampling methodology and outright lying by those polled. Political polling experts discuss proper polling methodology, and explain how pollsters can avoid dishonest responses that skew results and predictions. Produced by the Open University. (25 minutes)